diff --git a/.DS_Store b/.DS_Store
deleted file mode 100644
index b0c8a765a5d9a446e18ead874510ac99e3bf575e..0000000000000000000000000000000000000000
Binary files a/.DS_Store and /dev/null differ
diff --git a/.gitignore b/.gitignore
index 96609fc475f1331772ef9da6004b552033f951d1..de0396c5428ec469e24ab6de53f76de7e3b66fc5 100644
--- a/.gitignore
+++ b/.gitignore
@@ -76,7 +76,7 @@ docs/_build/
 target/
 
 # Jupyter Notebook
-.ipynb_checkpoints
+**/.ipynb_checkpoints
 
 # IPython
 profile_default/
@@ -160,3 +160,4 @@ cython_debug/
 #.idea/
 
 *.bak
+.DS_Store
diff --git a/BLcourse1/colab/.DS_Store b/BLcourse1/colab/.DS_Store
deleted file mode 100644
index 5008ddfcf53c02e82d7eee2e57c38e5672ef89f6..0000000000000000000000000000000000000000
Binary files a/BLcourse1/colab/.DS_Store and /dev/null differ
diff --git a/BLcourse3/.ipynb_checkpoints/svb_biexp_tf2-checkpoint.ipynb b/BLcourse3/.ipynb_checkpoints/svb_biexp_tf2-checkpoint.ipynb
deleted file mode 100644
index b448b5171d3b011d7eefbcd2ebf405bf091a10b2..0000000000000000000000000000000000000000
--- a/BLcourse3/.ipynb_checkpoints/svb_biexp_tf2-checkpoint.ipynb
+++ /dev/null
@@ -1,540 +0,0 @@
-{
- "cells": [
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "Stochastic Variational Bayes - example nonlinear model\n",
-    "==============================================\n",
-    "\n",
-    "This notebook implements stochastic variational Bayes for a nonlinear model. The model we will use is a bi-exponential model, i.e. we will assume our data reflects a time-dependent signal of the following form:\n",
-    "\n",
-    "$$S_{true}(t) = A_1 e^{-R_1t} + A_2 e^{-R_2t}$$\n",
-    "\n",
-    "However the actual time dependent signal $S(t)$ will be affected by additive Gaussian noise, so will have the distribution:\n",
-    "\n",
-    "$$P(S(t)) = \\frac{\\sqrt{\\beta}}{\\sqrt{2\\pi}} \\exp{\\bigg(-\\frac{\\beta}{2} (S(t) - S_{true}(t))^2}\\bigg)$$\n",
-    "\n",
-    "Given $S(t)$ our aim will be to infer the values of $A_1$, $A_2$, $R_1$, $R_2$ and $\\beta$.\n",
-    "\n",
-    "Here's how we can generate some sample data from this model in Python:"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 1,
-   "metadata": {
-    "vscode": {
-     "languageId": "python"
-    }
-   },
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Data samples are:\n",
-      "[ 2.03620472e+01  1.76781216e+01  1.39776282e+01  1.10712623e+01\n",
-      "  1.16523916e+01  1.30277908e+01  1.25537193e+01  1.01704583e+01\n",
-      "  7.96354263e+00  9.20085070e+00  8.75754246e+00  9.35617879e+00\n",
-      "  7.89357752e+00  7.53311329e+00  5.91293872e+00  5.24376529e+00\n",
-      "  6.28336297e+00  6.71847155e+00  5.71790066e+00  5.25259236e+00\n",
-      "  6.56513068e+00  6.95087490e+00  6.12893516e+00  7.12010747e+00\n",
-      "  6.39054434e+00  3.36110237e+00  5.24067202e+00  5.17919296e+00\n",
-      "  4.53724675e+00  4.20705702e+00  4.31729696e+00  2.78275598e+00\n",
-      "  5.56552827e+00  5.43811444e+00  3.58719833e+00  4.85273364e+00\n",
-      "  4.53756383e+00  2.42757610e+00  3.13918933e+00  2.63059348e+00\n",
-      "  2.03080086e+00  2.53277044e+00  2.79351027e+00  4.18524491e+00\n",
-      "  2.84583052e+00  2.47523917e+00  2.80880143e+00  2.88411467e+00\n",
-      "  4.35590416e+00  2.38959304e+00  3.70478916e+00  1.26115357e+00\n",
-      "  3.53994629e+00  2.26371690e+00  3.67565582e+00  1.81290846e+00\n",
-      "  4.87808803e-01  1.71551353e+00  2.81804463e+00  1.65500374e+00\n",
-      "  2.05152381e+00  2.31362865e+00  2.64819312e+00  1.04163689e-01\n",
-      "  2.15565841e+00  2.84082623e+00  2.22363156e+00  1.69315977e+00\n",
-      " -1.48598226e-02  9.32533929e-03  2.21885990e+00  1.72331807e+00\n",
-      "  8.73697296e-01  1.58991128e+00  3.45045992e+00  9.90859664e-01\n",
-      "  1.15792744e+00  1.02745892e+00  1.97796970e+00  1.50277410e+00\n",
-      "  2.44121134e+00  3.18994846e+00  1.93163364e+00  1.06704942e+00\n",
-      "  1.55037340e+00  1.00462960e+00  1.35572740e+00  1.75299392e-01\n",
-      "  1.58040422e+00  2.48303428e+00  1.14703694e+00  8.54837489e-01\n",
-      "  1.38005522e+00  2.52592716e-01 -4.30831092e-01  1.49594355e+00\n",
-      "  1.47165632e+00  1.68166865e+00 -1.01012022e+00  1.67268457e+00\n",
-      "  1.64854603e+00 -5.14833991e-01  3.52822298e-01 -2.28589149e-01\n",
-      "  6.29631984e-01  1.39003582e+00  1.79227013e+00 -1.45065940e-02\n",
-      "  7.84766104e-01  1.76306832e+00 -4.90317522e-01  6.50280679e-01\n",
-      " -2.57597827e+00  9.67710157e-01  1.20854107e-01  8.04335447e-01\n",
-      "  1.41536283e+00 -3.08961604e-01  1.36912774e+00  4.39539872e-01\n",
-      " -2.99185860e-01  8.24312201e-01  4.49438227e-01 -5.59622708e-01\n",
-      " -1.94886082e-01  1.41361978e+00  3.53411754e-01 -2.15636575e+00\n",
-      "  6.42295292e-01 -1.10171873e-01  4.52087704e-01  2.50551787e-01\n",
-      " -2.76845582e-01 -3.40397981e-02 -1.52621755e+00  2.21713254e-01\n",
-      " -8.94173404e-01  3.84562605e-01 -2.16465922e-01  1.26362820e+00\n",
-      " -8.64662292e-01 -1.80149764e-01 -2.50034817e-01  4.40519169e-01\n",
-      "  8.14419526e-01  9.38449237e-01  1.19993681e+00 -1.39000596e+00\n",
-      " -3.43890752e-01  1.30759870e-01  2.70952957e-01  1.37185684e-01\n",
-      "  1.54659280e+00  9.64976063e-01 -3.46476770e-02  2.09807350e+00\n",
-      " -2.12499002e-02 -1.03933571e+00 -1.31959675e+00  9.99991576e-01\n",
-      " -1.24689743e+00  1.66216228e+00  1.07757694e-02  9.33931324e-01\n",
-      " -4.23189470e-01  1.33272005e+00  6.06856755e-01  3.03137652e-01\n",
-      "  1.91326183e-01 -6.27514905e-01 -4.17695667e-01 -1.22337127e+00\n",
-      " -4.74243936e-03  6.17601284e-01 -6.06842773e-01  1.18880575e-01\n",
-      "  1.05808811e+00 -1.25474748e-01  1.56192277e+00 -1.49168252e+00\n",
-      "  8.24082426e-01 -6.95631429e-01  8.49622698e-02 -1.34546320e+00\n",
-      "  5.71835405e-01  8.10534452e-01 -1.31301694e-01 -3.72783179e-01\n",
-      " -3.74598991e-01 -6.93093569e-01  1.15441219e+00 -3.57336625e-01\n",
-      " -5.39608652e-01  2.00796488e+00  1.48405142e+00  1.05920302e+00\n",
-      " -7.56504039e-01  1.17203209e+00  4.39541412e-01 -7.85448587e-01]\n"
-     ]
-    }
-   ],
-   "source": [
-    "import numpy as np\n",
-    "%matplotlib inline\n",
-    "# Ground truth parameters\n",
-    "# We infer the precision, BETA, but it is useful to\n",
-    "# derive the variance and standard deviation from it\n",
-    "A1_TRUTH = 10.0\n",
-    "A2_TRUTH = 10.0\n",
-    "R1_TRUTH = 10.0\n",
-    "R2_TRUTH = 1.0\n",
-    "BETA_TRUTH = 1.0\n",
-    "VAR_TRUTH = 1/BETA_TRUTH\n",
-    "STD_TRUTH = np.sqrt(VAR_TRUTH)\n",
-    "\n",
-    "# Observed data samples are generated by Numpy from the ground truth\n",
-    "# Gaussian distribution. Reducing the number of samples should make\n",
-    "# the inference less 'confident' - i.e. the output variances for\n",
-    "# MU and BETA will increase\n",
-    "N = 200\n",
-    "T = np.linspace(0, 5, N)\n",
-    "\n",
-    "\n",
-    "DATA_CLEAN = A1_TRUTH * np.exp(-R1_TRUTH * T) + A2_TRUTH * np.exp(-R2_TRUTH*T)\n",
-    "DATA = DATA_CLEAN + np.random.normal(0, STD_TRUTH, [N])\n",
-    "print(\"Data samples are:\")\n",
-    "print(DATA)"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "We can plot this data to illustrate the true signal (green line) and the measured data (red crosses):"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 2,
-   "metadata": {
-    "vscode": {
-     "languageId": "python"
-    }
-   },
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "[<matplotlib.lines.Line2D at 0x14bca86cbbb0>]"
-      ]
-     },
-     "execution_count": 2,
-     "metadata": {},
-     "output_type": "execute_result"
-    },
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 432x288 with 1 Axes>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "from matplotlib import pyplot as plt\n",
-    "plt.figure()\n",
-    "plt.plot(DATA, \"rx\")\n",
-    "plt.plot(DATA_CLEAN, \"g\")"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "As with the single Gaussian example we will use a multivariate normal distribution as our prior and approximate posterior distributions. \n",
-    "\n",
-    "One difference here is that we will choose to infer the log of the decay rate parameters $R_1$ and $R_2$. This is because if these parameters are allowed to become negative the model prediction will become an exponential growth and can easily lead to numerical errors.\n",
-    "\n",
-    "We will still choose our priors to be relatively uninformative as follows:\n"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 3,
-   "metadata": {
-    "vscode": {
-     "languageId": "python"
-    }
-   },
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Priors: Amplitude mean=1.000000, variance=100000.000000\n",
-      "        Log decay rate mean=0.000000, variance=10.000000\n",
-      "        Log noise variance mean=0.000000, variance=10.000000\n"
-     ]
-    }
-   ],
-   "source": [
-    "a0 = 1.0\n",
-    "v0 = 100000.0\n",
-    "r0 = 0.0\n",
-    "u0 = 10.0\n",
-    "b0 = 0.0\n",
-    "w0 = 10.0\n",
-    "print(\"Priors: Amplitude mean=%f, variance=%f\" % (a0, v0))\n",
-    "print(\"        Log decay rate mean=%f, variance=%f\" % (r0, u0))\n",
-    "print(\"        Log noise variance mean=%f, variance=%f\" % (b0, w0))"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "The posterior will be defined in the same way as for the single Gaussian example, however we need to account for the increased number of parameters we are inferring. We will initialize the posterior with the prior values but with reduced initial variance to prevent problems with generating a representative posterior sample. Remember that the decay rates (and the noise) are being inferred as their log-values so the prior mean of 0 translates into a value of 1.\n",
-    "\n",
-    "The code to generate a posterior sample is unchanged except for the number of parameters.\n",
-    "\n",
-    "In calculating the reconstruction cost we need to calculate the log likelihood of the data given a set of model parameters. We do this by observing that any difference between the biexponential model prediction (given these parameters) and the actual noisy data must be a result of the Gaussian noise - hence the likelihood is simply the likelihood of drawing these differences from the Gaussian noise distribution:\n",
-    "\n",
-    "$$\\log P(\\textbf{y} | A_1; A_2; r_1; r_2; \\beta) = \\frac{1}{2} \\bigg( N \\log \\beta - \\sum{\\frac{(y_n - M_n)^2}{\\beta}}\\bigg)$$\n",
-    "\n",
-    "Here $M_n$ is the model prediction for the nth data point which is calculated by evaluating the biexponential model for the given parameters $A_1$, $A_2$, $r_1$ and $r_2$.\n",
-    "\n",
-    "For the latent loss we will again use the analytic expression for the K-L divergence of two MVN distributions with a slight modification to the previous code to account for the different number of parameters (5 vs 2)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 4,
-   "metadata": {
-    "vscode": {
-     "languageId": "python"
-    }
-   },
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "2023-03-19 17:07:51.945965: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1510] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 38284 MB memory:  -> device: 0, name: NVIDIA A100-SXM4-40GB, pci bus id: 0000:03:00.0, compute capability: 8.0\n",
-      "2023-03-19 17:07:51.947931: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1510] Created device /job:localhost/replica:0/task:0/device:GPU:1 with 38284 MB memory:  -> device: 1, name: NVIDIA A100-SXM4-40GB, pci bus id: 0000:44:00.0, compute capability: 8.0\n",
-      "2023-03-19 17:07:51.949520: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1510] Created device /job:localhost/replica:0/task:0/device:GPU:2 with 38284 MB memory:  -> device: 2, name: NVIDIA A100-SXM4-40GB, pci bus id: 0000:84:00.0, compute capability: 8.0\n",
-      "2023-03-19 17:07:51.951196: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1510] Created device /job:localhost/replica:0/task:0/device:GPU:3 with 38284 MB memory:  -> device: 3, name: NVIDIA A100-SXM4-40GB, pci bus id: 0000:c4:00.0, compute capability: 8.0\n"
-     ]
-    }
-   ],
-   "source": [
-    "import tensorflow as tf\n",
-    "\n",
-    "# Number of parameters - 4 for the biexponential + noise\n",
-    "NUM_PARAMS = 4 + 1\n",
-    "\n",
-    "data = tf.constant(DATA, dtype=tf.float32)\n",
-    "prior_means = tf.constant([a0, r0, a0, r0, b0], dtype=tf.float32)\n",
-    "prior_covariance = tf.linalg.diag(tf.constant([v0, u0, v0, u0, w0], dtype=tf.float32))\n",
-    "\n",
-    "post_means_init = prior_means\n",
-    "post_covariance_init = np.identity(NUM_PARAMS, dtype=np.float32)\n",
-    "\n",
-    "chol_off_diag = tf.Variable(np.zeros(post_covariance_init.shape), dtype=tf.float32)\n",
-    "# Comment in this line if you do NOT want to infer parameter covariances\n",
-    "#chol_off_diag = tf.constant([[0, 0], [0, 0]], dtype=tf.float32)\n",
-    "chol_log_diag = tf.Variable(tf.math.log(tf.linalg.diag_part(post_covariance_init)), dtype=tf.float32)\n",
-    "\n",
-    "post_means = tf.Variable(post_means_init, dtype=tf.float32)\n",
-    "\n",
-    "t = tf.reshape(tf.constant(T, dtype=tf.float32), [1, -1])\n",
-    "\n",
-    "\n",
-    "\n",
-    "def cost_fun():\n",
-    "\n",
-    "    chol_diag = tf.linalg.diag(tf.math.sqrt(tf.math.exp(chol_log_diag)))\n",
-    "    post_covariance_chol = tf.math.add(chol_diag, tf.linalg.band_part(chol_off_diag, -1, 0))\n",
-    "\n",
-    "    post_covariance = tf.matmul(tf.transpose(post_covariance_chol), post_covariance_chol)\n",
-    "\n",
-    "    S=5\n",
-    "    N=200\n",
-    "\n",
-    "# eps is a sample from a Gaussian with mean 0 and variance 1\n",
-    "    eps = tf.random.normal((NUM_PARAMS, S), 0, 1, dtype=tf.float32)\n",
-    "\n",
-    "# Start off each sample with the current posterior mean\n",
-    "# post_samples is now a tensor of shape [NUM_PARAMS, n_samples]\n",
-    "    samples = tf.tile(tf.reshape(post_means, [NUM_PARAMS, 1]), [1, S])\n",
-    "\n",
-    "# Now add the random sample scaled by the covariance\n",
-    "    post_samples = tf.add(samples, tf.matmul(post_covariance_chol, eps))\n",
-    "\n",
-    "    a1 = tf.reshape(post_samples[0], [-1, 1])\n",
-    "    r1 = tf.math.exp(tf.reshape(post_samples[1], [-1, 1]))\n",
-    "    a2 = tf.reshape(post_samples[2], [-1, 1])\n",
-    "    r2 = tf.math.exp(tf.reshape(post_samples[3], [-1, 1]))\n",
-    "\n",
-    "# Get the current estimate of the noise variance remembering that\n",
-    "# we are inferring the log of the noise precision, beta\n",
-    "    log_noise_var = -post_samples[4]\n",
-    "    noise_var = tf.math.exp(log_noise_var)\n",
-    "\n",
-    "# Each sample value predicts the full set of values in the data sample.\n",
-    "# For our constant-signal model, the prediction is simply a set of \n",
-    "# constant values. The prediction tensor will have shape [S, N]\n",
-    "# where S is the sample size and N is the number of data values\n",
-    "    \n",
-    "\n",
-    "    prediction = a1*tf.math.exp(-r1*t) + a2*tf.exp(-r2*t)\n",
-    "    diff = tf.reshape(data, [1, -1]) - prediction\n",
-    "\n",
-    "# To calculate the likelihood we need the sum of the squared difference between the data  \n",
-    "# and the prediction. This gives a value for each posterior sample so has shape [S]\n",
-    "    sum_square_diff = tf.reduce_sum(tf.math.square(diff), axis=1)\n",
-    "\n",
-    "# Now we calculate the likelihood for each posterior sample (shape [S])\n",
-    "# Note that we are ignoring constant factors such as 2*PI here as they \n",
-    "# are just an fixed offset and do not affect the optimization \n",
-    "    log_likelihood = 0.5 * (-log_noise_var * tf.cast(N,dtype=tf.float32) - sum_square_diff / noise_var)\n",
-    "\n",
-    "# Finally to evaluate the expectation value we take the mean across all the posterior\n",
-    "# samples. The negative of this is the reconstruction loss\n",
-    "    reconstr_loss = -tf.reduce_mean(log_likelihood)\n",
-    "\n",
-    "    C = post_covariance\n",
-    "    C0 = prior_covariance\n",
-    "    #print(\"is this the problem? 1\")\n",
-    "    #print(tf.linalg.det(C0))\n",
-    "    C0_inv = tf.linalg.inv(C0)\n",
-    "    #print(\"is this the problem? 2\")\n",
-    "\n",
-    "# m - m0 as row and column vectors\n",
-    "    m_minus_m0 = tf.reshape(tf.subtract(post_means, prior_means), [-1, 1])\n",
-    "    m_minus_m0_T = tf.reshape(tf.subtract(post_means, prior_means), [1, -1])\n",
-    "\n",
-    "    term1 = tf.linalg.trace(tf.matmul(C0_inv, C))\n",
-    "    term2 = -tf.math.log(tf.linalg.det(C) / tf.linalg.det(C0))\n",
-    "\n",
-    "# Size of the MVN distribution\n",
-    "    term3 = -NUM_PARAMS\n",
-    "    term4 = tf.matmul(tf.matmul(m_minus_m0_T, C0_inv), m_minus_m0)\n",
-    "          \n",
-    "    latent_loss = 0.5 * (term1 + term2 + term3 + term4)\n",
-    "\n",
-    "    #cost = reconstr_loss + latent_loss\n",
-    "    cost=reconstr_loss\n",
-    "\n",
-    "    return cost\n",
-    "\n",
-    "\n"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "Finally we ask TensorFlow to minimise the total cost iteratively:"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 7,
-   "metadata": {
-    "vscode": {
-     "languageId": "python"
-    }
-   },
-   "outputs": [
-    {
-     "data": {
-      "application/vnd.jupyter.widget-view+json": {
-       "model_id": "cd06775d27544b8b8190d373e6d0b97a",
-       "version_major": 2,
-       "version_minor": 0
-      },
-      "text/plain": [
-       "Epoch::   0%|          | 0/5000 [00:00<?, ?it/s]"
-      ]
-     },
-     "metadata": {},
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "optimizer = tf.keras.optimizers.Adam(learning_rate=0.02)\n",
-    "#minimizer = optimizer.minimize(cost)\n",
-    "#sess.run(tf.global_variables_initializer())\n",
-    "from tqdm.notebook import tqdm_notebook\n",
-    "\n",
-    "cost_history = []\n",
-    "for epoch in tqdm_notebook(range(5000), desc=\"Epoch:\"):\n",
-    "    #sess.run(minimizer)\n",
-    "\n",
-    "    optimizer.minimize(cost_fun,var_list=[chol_off_diag,chol_log_diag,post_means])\n",
-    "    #print(float(cost_fun()))\n",
-    "    cost_history.append(float(cost_fun()))\n",
-    "    #print(\"Epoch %i: posterior means=%s\" % (epoch+1, post_means))"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 8,
-   "metadata": {
-    "vscode": {
-     "languageId": "python"
-    }
-   },
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Estimate for amp1: 10.243862 (variance: 0.000000)\n",
-      "Estimate for amp2: 9.581695 (variance: 0.000000)\n",
-      "Estimate for r1: 1.075554\n",
-      "Estimate for r2: 11.576426\n",
-      "Estimate for beta (noise): 0.912233\n"
-     ]
-    }
-   ],
-   "source": [
-    "final_means = post_means\n",
-    "\n",
-    "chol_diag = tf.linalg.diag(tf.math.sqrt(tf.math.exp(chol_log_diag)))\n",
-    "post_covariance_chol = tf.add(chol_diag, tf.linalg.band_part(chol_off_diag, -1, 0))\n",
-    "\n",
-    "post_covariance = tf.matmul(tf.transpose(post_covariance_chol), post_covariance_chol)\n",
-    "\n",
-    "final_covariance = post_covariance\n",
-    "print(\"Estimate for amp1: %f (variance: %f)\" % (final_means[0], final_covariance[0, 0]))\n",
-    "print(\"Estimate for amp2: %f (variance: %f)\" % (final_means[2], final_covariance[0, 0]))\n",
-    "print(\"Estimate for r1: %f\" % (np.exp(final_means[1]),))\n",
-    "print(\"Estimate for r2: %f\" % (np.exp(final_means[3]),))\n",
-    "print(\"Estimate for beta (noise): %f\" % np.exp(-final_means[4]))"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 10,
-   "metadata": {
-    "vscode": {
-     "languageId": "python"
-    }
-   },
-   "outputs": [
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 432x288 with 1 Axes>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    },
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 432x288 with 1 Axes>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "plt.figure()\n",
-    "plt.plot(cost_history)\n",
-    "plt.ylim(0, 500)\n",
-    "\n",
-    "plt.figure()\n",
-    "plt.plot(DATA_CLEAN, 'g')\n",
-    "plt.plot(DATA, 'rx')\n",
-    "S=5\n",
-    "eps = tf.random.normal((NUM_PARAMS, S), 0, 1, dtype=tf.float32)\n",
-    "\n",
-    "samples = tf.tile(tf.reshape(post_means, [NUM_PARAMS, 1]), [1, S])\n",
-    "\n",
-    "# Now add the random sample scaled by the covariance\n",
-    "post_samples = tf.add(samples, tf.matmul(post_covariance_chol, eps))\n",
-    "\n",
-    "a1 = tf.reshape(post_samples[0], [-1, 1])\n",
-    "r1 = tf.math.exp(tf.reshape(post_samples[1], [-1, 1]))\n",
-    "a2 = tf.reshape(post_samples[2], [-1, 1])\n",
-    "r2 = tf.math.exp(tf.reshape(post_samples[3], [-1, 1]))\n",
-    "\n",
-    "# Get the current estimate of the noise variance remembering that\n",
-    "# we are inferring the log of the noise precision, beta\n",
-    "log_noise_var = -post_samples[4]\n",
-    "noise_var = tf.math.exp(log_noise_var)\n",
-    "\n",
-    "# Each sample value predicts the full set of values in the data sample.\n",
-    "# For our constant-signal model, the prediction is simply a set of \n",
-    "# constant values. The prediction tensor will have shape [S, N]\n",
-    "# where S is the sample size and N is the number of data values\n",
-    "t = tf.reshape(tf.constant(T, dtype=tf.float32), [1, -1])\n",
-    "prediction = a1*tf.math.exp(-r1*t) + a2*tf.exp(-r2*t)\n",
-    "\n",
-    "plt.plot(prediction[0], 'b')\n",
-    "plt.show()"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {
-    "vscode": {
-     "languageId": "python"
-    }
-   },
-   "outputs": [],
-   "source": []
-  }
- ],
- "metadata": {
-  "kernelspec": {
-   "display_name": "PyDeepLearning-1.1",
-   "language": "python",
-   "name": "pydeeplearning"
-  },
-  "language_info": {
-   "codemirror_mode": {
-    "name": "ipython",
-    "version": 3
-   },
-   "file_extension": ".py",
-   "mimetype": "text/x-python",
-   "name": "python",
-   "nbconvert_exporter": "python",
-   "pygments_lexer": "ipython3",
-   "version": "3.9.6"
-  },
-  "vscode": {
-   "interpreter": {
-    "hash": "31f2aee4e71d21fbe5cf8b01ff0e069b9275f58929596ceb00d14d90e3e16cd6"
-   }
-  }
- },
- "nbformat": 4,
- "nbformat_minor": 4
-}
diff --git a/BLcourse3/.ipynb_checkpoints/svb_gaussian_tf2-checkpoint.ipynb b/BLcourse3/.ipynb_checkpoints/svb_gaussian_tf2-checkpoint.ipynb
deleted file mode 100644
index 53dc2789cc116ab00f33561813c199a8698db619..0000000000000000000000000000000000000000
--- a/BLcourse3/.ipynb_checkpoints/svb_gaussian_tf2-checkpoint.ipynb
+++ /dev/null
@@ -1,571 +0,0 @@
-{
- "cells": [
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "Stochastic Variational Bayes\n",
-    "=======================\n",
-    "\n",
-    "This notebook implements Example 1 from the FMRIB tutorial on Variational Bayes II: Stochastic Variational Bayes (\"fitting a Gaussian distribution).\n",
-    "\n",
-    "We assume we have data drawn from a Gaussian distribution with true mean $\\mu$ and true precision $\\beta$:\n",
-    "\n",
-    "$$\n",
-    "P(y_n | \\mu, \\beta) = \\frac{\\sqrt{\\beta}}{\\sqrt{2\\pi}} \\exp{-\\frac{\\beta}{2} (y_n - \\mu)^2}\n",
-    "$$\n",
-    "\n",
-    "One interpretation of this is that our data consists of repeated measurements of a fixed value ($\\mu$) combined with Gaussian noise with standard deviation $\\frac{1}{\\sqrt{\\beta}}$.\n",
-    "\n",
-    "Here's how we can generate some sample data from this model in Python:"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 1,
-   "metadata": {
-    "vscode": {
-     "languageId": "python"
-    }
-   },
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Data samples are:\n",
-      "[41.69330425 43.22760142 42.04227699 41.8726123  42.10124698 42.06148383\n",
-      " 43.07223824 42.95316759 41.38239292 42.80939571 44.77113522 40.59558767\n",
-      " 42.94639899 42.60768533 41.44394187 42.78759534 39.98116589 42.5661791\n",
-      " 42.48842924 40.99124449 41.89709575 43.11330919 41.708735   42.87831323\n",
-      " 41.08147171 41.71958743 42.46591807 41.6370583  41.8648694  43.29941234\n",
-      " 41.53153621 42.27939892 42.73532511 42.80090713 42.41404559 42.16952383\n",
-      " 40.97795771 43.72456543 42.93386718 42.75672866 41.39694193 39.7078651\n",
-      " 43.91464577 43.33962568 40.09870405 43.33660874 40.91621448 40.6448031\n",
-      " 43.04706312 42.12147857 41.38607984 42.21791959 41.12890543 41.97966952\n",
-      " 41.00969587 42.84116868 41.5972026  41.2236852  41.58269978 42.08416207\n",
-      " 41.91320553 39.37739812 41.77886506 41.78896676 41.96599989 40.52331769\n",
-      " 41.99641274 40.43126556 42.61890785 41.12090487 43.18604389 40.98318213\n",
-      " 43.23229514 44.00299414 41.38041941 43.05466329 42.01280297 40.38765589\n",
-      " 43.44037223 42.01792412 42.14212663 40.49368043 42.13380001 41.58396877\n",
-      " 41.00460675 43.25897542 40.55767639 41.25966307 41.61961251 41.99744515\n",
-      " 41.5689065  40.32893455 40.63016246 42.22022195 42.37517688 40.98311719\n",
-      " 41.13315559 41.54570082 42.34490129 40.07093031]\n"
-     ]
-    }
-   ],
-   "source": [
-    "import numpy as np\n",
-    "%matplotlib inline\n",
-    "\n",
-    "# Ground truth parameters\n",
-    "# We infer the precision, BETA, but it is useful to\n",
-    "# derive the variance and standard deviation from it\n",
-    "MU_TRUTH = 42\n",
-    "BETA_TRUTH = 1.0\n",
-    "VAR_TRUTH = 1/BETA_TRUTH\n",
-    "STD_TRUTH = np.sqrt(VAR_TRUTH)\n",
-    "\n",
-    "# Observed data samples are generated by Numpy from the ground truth\n",
-    "# Gaussian distribution. Reducing the number of samples should make\n",
-    "# the inference less 'confident' - i.e. the output variances for\n",
-    "# MU and BETA will increase\n",
-    "N = 100\n",
-    "DATA = np.random.normal(MU_TRUTH, STD_TRUTH, [N])\n",
-    "print(\"Data samples are:\")\n",
-    "print(DATA)"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "In the 'signal + noise' interpretation we can view this as noisy measurements (red crosses) of a constant signal (green line):"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 2,
-   "metadata": {
-    "vscode": {
-     "languageId": "python"
-    }
-   },
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "[<matplotlib.lines.Line2D at 0x1458624f6c10>]"
-      ]
-     },
-     "execution_count": 2,
-     "metadata": {},
-     "output_type": "execute_result"
-    },
-    {
-     "data": {
-      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXAAAAD7CAYAAABzGc+QAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAAAY40lEQVR4nO3da6wcZ33H8e+/TgsNxTE0BkIc56RVRIEiAjlGtFRVTlxoLlZCVNVK1bh5UelU7oXQgCyi6FjFEQpxSQkviiuaXiICpFbKJfULCopP1DcV9TG5KBdo0sZOHBJiWtyiInFp/n0xu/V6vbM7uzvPzPM88/tIqz07eznzzOU3z/7nsubuiIhIen6i7REQEZHZKMBFRBKlABcRSZQCXEQkUQpwEZFEKcBFRBJVOcDNbJ2ZPWhmB3qP/8TMnjOzh3q3K8KNpoiIDDtjitfeADwBrB8Y9nF3/1i9oyQiIlVUCnAz2wRcCXwEuHHWf3b22Wf7wsLCrG8XEemkw4cPf8fdNw4Pr9oDvwPYBbxyaPgfmtnvAGvAB9z9u+M+ZGFhgbW1tYr/UkREAMzs6KjhE2vgZrYNeNHdDw89tQ/4eeAi4Hng9pL3L5vZmpmtHT9+fKqRFhGRclV2Yr4LuMrMjgD3AJea2d3u/m13/193fwn4S+Ado97s7p9y90V3X9y48bRvACIiMqOJAe7uN7n7JndfAK4FDrr7dWZ2zsDLrgEeDTSOIiIywjRHoQzba2YXAQ4cAX6vjhESEZFqpgpwd38AeKD3944A4yMiIhXpTMwye/fC6uqpw1ZXi+EiIhFQgJfZsgW2bz8Z4qurxeMtW9odLxGRnnlq4HlbWoL9+4vQ3rkT9u0rHi8ttT1mIiKAeuDjLS0V4X3LLcW9wltEIqIAH2d1teh5r6wU98M1cRGRFinAy/Rr3vv3w549J8spCnERiYQCvMyhQ6fWvPs18UOH2h0vEZEec/fG/tni4qLrYlYiItMxs8Puvjg8XD1wEZFEKcBFRBKlABcRSZQCXEQkUQpwEZFEKcBFRBKlABcRSZQCXEQkUQpwEZFEKcBFRBKlABcRSZQCXEQkUQpwEZFEKcBFRBKlABcRSZQCXEQkUQpwEZFEKcBFRBKlABcRSZQCXEQkUZUD3MzWmdmDZnZgaPgHzczN7Oz6R09EWrV3L6yunjpsdbUYLq2bpgd+A/DE4AAzOw94N/BMnSMlIpHYsgW2bz8Z4qurxeMtW9odLwEqBriZbQKuBO4ceurjwC7Aax4vEYnB0hLs31+E9u7dxf3+/cVwaV3VHvgdFEH9Un+AmV0FPOfuDwcYLxGJxdIS7NwJt9xS3Cu8ozExwM1sG/Ciux8eGHYmcDOwu8L7l81szczWjh8/PtfISgepBtu+1VXYtw9WVor74fkhranSA38XcJWZHQHuAS4FPg1cADzcG74J+LqZvW74ze7+KXdfdPfFjRs31jbi0hGqwbarP73374c9e06WUxTiUZgY4O5+k7tvcvcF4FrgoLv/hru/xt0XesOPAW939xfCjq50jmqw7Tp06NTp3Z8fhw61O14CwBltj4DIRIM12JUVhXeTdu06fdjSkuZBJKY6kcfdH3D3bSOGL7j7d+obLZEBqsGKjKQzMSVuqsGKlFKAS9xUgxUpZe7NnYOzuLjoa2trjf0/icTevcVRI4N109XVIoRH1VhF5BRmdtjdF4eHqwcu4elQQJEgdBSKhDd4KODOncWOSB0KKDI39cClGTodW6R2CnBphg4FFKmdAlzC06GAIkEowCU8HQooEoQOIxQRiZwOIxQRyYwCXEQkUQpwEQlDP8YRnAJcRMLQGbjB6UxMEQlDZ+AGpx64iISjM3CDUoBLOsbVVFVvjZPOwA1KAd51KQXfuJqq6q3x0Rm4wSnAuy6l4Bv3A8f68eP46Azc8Ny9sdvFF1/sEqGDB93PPtt9ZaW4P3iw7TEab2XFHYr7aZ6T9N122+nL58GDxfCMAWs+IlPT6YGn9FU/NSntaBpXU1W9tR4xr2spfWNswqhUD3Wbqwfe7yX2t77Dj2V2qfTAxy0DWj6mM64nG/u0TGV5rRElPfB0Aty9kzMuuNhX1kHjQqejX61nNmm+x76udaxUlkeAu3duxgWn4OuuSSEd67oW+8YlgDwCvIMzTgErQZWFdKzrWkrfGGuUfoB3dMZ1tt0SXllIx7zMdbRDk36AzzvjUp7xsfaGUpPyMlC3cSFdZTppWjZq7gAH1gEPAgd6j28BHgEeAr4CvH7SZ7R6HHjMvYoqYq1HpiT1ZaBO8wawpmWj6gjwG4HPDgT4+oHn3gf8xaTPaP1EnlR7sqmOd4w0LeujadmYuQIc2ATcD1zaD/Ch528C9k36nNYD3D29nqx6OvVLbRmImaZlI8oCvOqZmHcAu4CXBgea2UfM7Fngt4Hdlc8eakv/TL2tW+ETnzj9LL4qZ5o1fZaaridRL52tWR9Ny/aNSnU/tXe9Dfhk7+9LKO+Bf7jk/cvAGrC2efPmBrdZQ4bP2jvrLPf160efxVf1c0Y9lvHa3PmleVcfTctGMWsJBbgVOAYcAV4Avg/cPfSa84FHJ31WqyWU4eDoh/jWrdMveKr9za7NFV9HTtRH07JRMwe4nxrU/98DBy4cGP5HwL2T3h9FDXzQPPW7tmt/Ka9A2gCKTKUswOe5GuFHzexRM3sEeA9wwxyf1bx56ncx1P5SvipbSlc/nCTmK/dJ/kaleqhbND3wsq/xy8uTe7Ux1f7a6snWdQxxDj3wmJYHyRbJn4lZp7IAWl6evDLGdpZaG6WceUIrx8DLaYMkUVKAV1XHythUSLUZHLP+78GNW//vwY1bKnX8YW3vE5GsKcCnUcfKGDpcm+7JjvpWsWPHfNMpl964euASmAK8qjpXxpC9sqaPQhkO19tvdzcrQnye6ZR6+OWyEZKoKcCrqHNlTD2YRum3aceOIrxvv/3U4bO2MeXyQ8qHc0oyFOBV1LUy5twr64ftjh2nDp81tHLc0InUTAHepFx7ZXWHbc4buljkuix2TFmAz3Mij5SdxAGnn5yytAS7djUzXiH0TxTavx/27CnuB08kmoUu1BVeyid8yWSjUj3ULbse+LQ9yJR7QymPe9epTJU8VEIJZJqVQyUDaUvKO4pFJZRgprmuR79EsH077N59siSR8rVAJH4xXLtHglCAz2valWMw8N/61tGfpwshSV1C7LuQaCjA5zHLyjEY+IcOwTXXhNvBpCvliXYU521UXSXULbsa+LQ79kbVwNevL35YIsQOJtXcRbKAdmJGoCzwt24Nt4Mp9BEIOjpFJLiyAFcJpUm7do3eYfnww+F2MIX+8QQdZyw5i70MOSrVQ9063wMf1kSJo4ljgHWcseQqknM9UAklQqHLD03WwHWcseQqgnM9FOBd1FR9Wj1wyd00HZQA64MCXKZTNfx1pIvkbpZArvkbaVmAayemjFZ156SOM5aczXuuR+gzX0eleqibeuCJ6WJpRIdFyjy/29pwDVw9cCkX+hDEGOmwSBlcBvqXgB5cBsZdGrrpb6SjUj3UTT3wxEzbA8+l99rFbx5yqsiWAbQTU6Yyy1fBlHZoTtrY6LDIcMqm/eWXx9UBiGgZ6EaA59IDjMGs0zKynkupcRubVNqQqrJpf/vt8XQAIlsGuhHgKfUAcxZRz2WsUStpE8uQOhrlARlDcEaYI3kF+LgVYHABOPPMYqs+6nUSRgwr4DSGNzZNhGuEAdGKsg192x2ACDewcwc4sA54EDjQe/ynwDeAR4AvABsmfUZtAT5pBegvADt2aEVpUmrB1ObGJrUN3ThVAm/4NQcPFpdR3ro1vh54hOoI8BuBzw4E+HuAM3p/3wbcNukzai2hVP0K1q+raYEIL8KeS6kYNjZ19jTbnPZVpuXwPoazziquhT9YtgpVA09puSwxV4ADm4D7gUv7AT70/DXAZyZ9Tu018OEVoGxB2rEjjZqsNGfalbruEJi1p1k2HsvL7W6QqrSn/5qtW0+G9+BzoY5CiWFjPaeyALfiufHM7F7gVuCVwAfdfdvQ8/8A/J273z3ucxYXF31tbW3i/xv2/i+/n4deeOjUgSdOwOOPwevPhW89B296M3zve/DKV8KGDSdfd+wYPP00nHfeydcNPi9SRX956y8/w4+b+qxx74XT14kml/UjR+DoUTj/fFhYmP01IYzKi4Zz4KLXXcQdl90x03vN7LC7Lw4Pn3gmppltA15098Mlz98M/Bj4TMnzy2a2ZmZrx48fn3K0SwwutAsLxf3jj50e3idOwDNH4S1vOfV1J07UMx7SHRs2nFx+jhyZPbyh6Gj03/vss8WwfgcEiuWzP3ya8diwoQioo0eL+yYD6sSJIhjPP7+4H7WOVXlNKG1Om5BGdcv91PLIrcAx4AjwAvB94O7ec9cD/wycOelzPNRRKOOuVZBq7SvV8e6Cuo+QmPXr/ajxaGsH4LQ18LLXNDGOie4Lo47DCIFLOLkT8zLgcWBj1fcHOQ687QUjhNBt0gZiNqFCYNrPHfX6NteDWY5CGfWaUDLIiBAB/hTwLPBQ7/YXk94f7ESexLeuI4VsUwYLdONCT7OqPfuy8Vhe1ka5TAYdlloCfN5b0DMx2z74P4SQbcpxoxdSyBCYZl5kEEYyvbwDPMcwaqJNOW70UlOlRz3tNaklO/kGeI7lgCbalONGL0VVjus+eHD0iS+aZ52Rb4Dn+JVynsttVpkeOW70cjS4kV2/vghxbXA7Kc0AzzGc51HX4VqarukYLHOp5NVZaQa4eoqnq1L6UHkkD+qBS0+aAe6uMBqlSk+sjcukSn0GOyuqgXdeWYDH/6PGXfxh3XFWV2HfPlhZKe77P7476TX6sd60DP447qFD8IUvwBe/WPwd+odyc7B37+nrxupqMTwno1I91E098DnNWwPXtAwrld96TMG83xgzK7+SZAkls5kwtzpOWdaOsHDKlteYfusxFXWs+6M6LCFKiQ2UJ9MMcNVt6zWpB67pPb+yaaxvP9OrY5pV/c2AeeZHAx3NNANc6lPXIYgyWay/9ZiieaZZkxvTwBtoBXjXVe1dq6c4H/XA6zPPNJvUGQmxMQ24gVaAh5ZT+UE9xdmoBl6feb8Njlsf1QNXgJ8ml/KDeoqz01Eo9QnVIVINPLMAr3NBST382twItfkNJqdvTzKejkLJLMDrDq2Uyw9d/dWUXL49haYNXWsU4OPU1XNOvQfetjann+bdZNrQtUYBPsm8PWct3PVo8xtMyt+emqINXSvKAjz+a6E0ocr1RSYZvHYF6HoVs6hjPqT4v1OiaxPFZVSqh7pF2QNXzzkOqoGnIeYeeMY1etQDL6GecxzanA8h/neOV8PrX8Fy/37Ys6e4H7zCZdu6eMXNUake6hZlD1wkhBx79Sn0cGP+hjAHSnrgVjzXjMXFRV9bW2vs/4m0qt8D3LmzqKsP9vIlnN27ixr9ykrxTSEDZnbY3ReHh6uEIhKKdvg1r2M7oxXgImXmrWPHFiY51uUHNVmjj2RaKsBzEMnClJ15dorFuMMv9518Te4Ij2VajiqMh7ppJ2YgOe4wi8WsO8Vi3eGX6U6+VjQ4LdGZmJnTihlObmdo5taeNjU0LcsCvHIJxczWmdmDZnag9/g3zewxM3vJzE7bOyoN0w6zMEbVsVMuWcVWl68i1ukdw7QcleqjbsCNwGeBA73HbwTeADwALFb5DPXAx6jrV7hj6YHHWkKYRllpKtUfaChrz/Jy3PMqlhLh4DI9uCwM/khEoHFinhIKsAm4H7i0H+ADzynA6zDPQhrLAh77OE2r6V91Ca2sPcvL8c+rGKb34HS57bbRG/JAG715A/xe4GLgkmkDHFgG1oC1zZs3B2lcNrTDLC051ZJTmFcxTO+WptPMAQ5sAz7Z+3vqAB+8qQdeQQwLaZ3aaE8TG7QUAm9aMS97MU3vFqbTPAF+K3AMOAK8AHwfuNsV4PWLaSGtQ1vtCV2+yaE8NCzmZS+m6Z1aD9xPDXP1wEOJaSGtQ9vtCbmixVqymlXb82qSWKZ3i9Op9gAHrun1zH8AfBv4x0nvV4CPEctCWpcY2hNzSSAmMcyrFLQ4ncoCXFcjlDzpSoCSEV2NULojxuuQiASgAJf86FeWpCNUQhERiZxKKCIimVGAS95ivRCSSA0U4JK3WC68L3FLdEOvAK8i0ZkrnNyBuX178WO3/aNTdEihDEp0Q68AryLRmSs9ula6TJLohl4BXkWiM1d6YrjwvoRR57fjBDf0CvCqEpy5Qt4n9ai0V++34xQ39KPOrw91S/paKDFfrU3K5Xydj9gvQlVFHfOnjnUz8mmJftR4DpHPXOmw1DsWda1b8164LPINvQJ8HpHPXOm44fBKbXmddyOU+kasAgW4SI5GhVeK3xhn7UGn2NYZlAW4dmKKpKpsBy20d9TULDtW59l52PULl41K9VA39cBFajSpVNLGD1pM2yPuSA96XqiEItIhbdaFp/nfbdbrE9pXoAAX6YoYerUp/JxdDNOporIAVw1cJDdt14VTOSEmgzOs9YMOIlKfwR2rS0unP47R7t3FGdYrK8XO4AjpBx1EJLy2e//TSuXbQgn1wEWkmxL6tqAeuIjIoNS+LYygHriISOTUA5e86FKqIgpwSZR+JUlSELijoQCXNGVwDK90QOCORuUAN7N1ZvagmR3oPX61mX3VzJ7s3b+qljESqUq/kiSxC9zRmKYHfgPwxMDjDwH3u/uFwP29xyLNSfwYXumIgB2NSgFuZpuAK4E7BwZfDdzV+/su4L21jVVOtLMtjJx/61LyErCjUbUHfgewC3hpYNhr3f15gN79a2obq5xoZ1sYGRzDKx0QuKMx8ThwM9sGXOHuv29mlwAfdPdtZnbC3TcMvO677n5aHdzMloFlgM2bN1989OjRWkY8Kf2ZuHNnsQXWzjaRbti7t+isDa7vq6tFR2PXrsofU3YceJUAvxXYAfwYeDmwHvg8sAW4xN2fN7NzgAfc/Q3jPqvTJ/IkcMEcEYnTzCfyuPtN7r7J3ReAa4GD7n4dcB9wfe9l1wNfqnF886KdbSISwDzHgX8UeLeZPQm8u/dYhmlnm4gEcsY0L3b3B4AHen//B7C1/lHKzLidbaqDi8gcdDErEZHI6WJWIiKZUYCLiCRKAS7x0dmrIpUowCU+OntVpJKpjkIRacTgFdx09qpIKfXAJU66VKzIRApwiZPOXhWZSAEu8dHZqyKVKMAlPrpUrEglOhNTRCRyOhNTRCQzCnARkUQpwEVEEqUAFxFJlAJcJBa6BoxMSQEuEgtdA0ampGuhiMRC14CRKakHLhITXQNGpqAAF4mJrgEjU1CAi8RC14CRKSnARWKha8DIlHQtFBGRyOlaKCIimVGAi4gkSgEuIpIoBbiISKIU4CIiiWr0KBQzOw4cnfHtZwPfqXF0UtHFdnexzdDNdnexzTB9u893943DAxsN8HmY2dqow2hy18V2d7HN0M12d7HNUF+7VUIREUmUAlxEJFEpBfin2h6BlnSx3V1sM3Sz3V1sM9TU7mRq4CIicqqUeuAiIjIgiQA3s8vM7Jtm9pSZfajt8QnBzM4zs1Uze8LMHjOzG3rDX21mXzWzJ3v3r2p7XOtmZuvM7EEzO9B73IU2bzCze83sG715/ku5t9vM/ri3bD9qZp8zs5fn2GYz+2sze9HMHh0YVtpOM7upl23fNLNfn+Z/RR/gZrYO+HPgcuBNwG+Z2ZvaHasgfgx8wN3fCLwT+INeOz8E3O/uFwL39x7n5gbgiYHHXWjzJ4Avu/svAG+laH+27Tazc4H3AYvu/ovAOuBa8mzz3wKXDQ0b2c7eOn4t8Obeez7Zy7xKog9w4B3AU+7+7+7+Q+Ae4OqWx6l27v68u3+99/f3KFbocynaelfvZXcB721lBAMxs03AlcCdA4Nzb/N64FeBvwJw9x+6+wkybzfFb/D+tJmdAZwJfIsM2+zu/wT859DgsnZeDdzj7j9w96eBpygyr5IUAvxc4NmBx8d6w7JlZgvA24CvAa919+ehCHngNS2OWgh3ALuAlwaG5d7mnwOOA3/TKx3daWavION2u/tzwMeAZ4Dngf9y96+QcZuHlLVzrnxLIcBtxLBsD50xs58B/h54v7v/d9vjE5KZbQNedPfDbY9Lw84A3g7sc/e3Af9DHqWDUr2a79XABcDrgVeY2XXtjlUU5sq3FAL8GHDewONNFF+9smNmP0kR3p9x98/3Bn/bzM7pPX8O8GJb4xfAu4CrzOwIRWnsUjO7m7zbDMUyfczdv9Z7fC9FoOfc7l8Dnnb34+7+I+DzwC+Td5sHlbVzrnxLIcAPARea2QVm9lMUBf/7Wh6n2pmZUdREn3D3Pxt46j7g+t7f1wNfanrcQnH3m9x9k7svUMzXg+5+HRm3GcDdXwCeNbM39AZtBR4n73Y/A7zTzM7sLetbKfbz5NzmQWXtvA+41sxeZmYXABcC/1L5U909+htwBfCvwL8BN7c9PoHa+CsUX50eAR7q3a4AfpZir/WTvftXtz2ugdp/CXCg93f2bQYuAtZ68/uLwKtybzfwYeAbwKPAp4GX5dhm4HMUdf4fUfSwf3dcO4Gbe9n2TeDyaf6XzsQUEUlUCiUUEREZQQEuIpIoBbiISKIU4CIiiVKAi4gkSgEuIpIoBbiISKIU4CIiifo/vb2679vsTpAAAAAASUVORK5CYII=\n",
-      "text/plain": [
-       "<Figure size 432x288 with 1 Axes>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "from matplotlib import pyplot as plt\n",
-    "plt.figure()\n",
-    "plt.plot(DATA, \"rx\")\n",
-    "plt.plot([MU_TRUTH] * N, \"g\")"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 3,
-   "metadata": {
-    "vscode": {
-     "languageId": "python"
-    }
-   },
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Priors: P(mu, log(1/beta)) = MVN([0.000000, 10000.000000], [[0.000000, 0], [0, 10000.000000]])\n"
-     ]
-    }
-   ],
-   "source": [
-    "m0 = 0.0\n",
-    "v0 = 10000.0\n",
-    "b0 = 0.0\n",
-    "w0 = 10000.0\n",
-    "print(\"Priors: P(mu, log(1/beta)) = MVN([%f, %f], [[%f, 0], [0, %f]])\" % (m0, v0, b0, w0))"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "As with analytic Variational Bayes, we need to choose an approximate form for our priors and posteriors. However we have more freedom in the stochastic method since we are not limited by the requirement that the distributions be conjugate with respect to the likelihood. \n",
-    "\n",
-    "We will choose a multivariate normal distribution (MVN) over the two parameters $\\mu$ and $log(\\frac{1}{\\beta})$ for prior and posterior. Inferring the log of the noise variance is useful as it avoid the possibility of negative values during the optimization which would make the likelihood ill-defined.\n",
-    "\n",
-    "The choice of an MVN means that we can allow for covariance (correlation) between the noise and signal parameters. This is unlike the analytic case where the posterior had to be factorised over these two parameters.\n",
-    "\n",
-    "An MVN distribution for $N$ parameters is defined by a vector of $N$ mean values and an $N \\times N$ covariance matrix. For the prior we will use the following values:\n",
-    "\n",
-    "$$\\textbf{m}_0 = \\begin{bmatrix} \\mu_0 \\\\ b_0 \\end{bmatrix}$$\n",
-    "\n",
-    "$$\\textbf{C}_0 = \\begin{bmatrix} v_0 & 0 \\\\ 0 & w_0 \\end{bmatrix}$$\n",
-    "\n",
-    "Note that we are not assuming any prior covariance. \n",
-    "\n",
-    "We define some suitable relatively uninformative prior values here:"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "Stochastic VB is based around minimising the free energy so we will need to implement the calculation of the free energy. We will be using the TensorFlow framework to perform the minimisation so the calculation must be in terms of tensors (multidimensional arrays). In our case the following constant tensors must be defined (where $N$ is the number of data values we have:\n",
-    "\n",
-    " - Data samples: $[N]$\n",
-    " - Prior mean: $[2]$\n",
-    " - Prior covariance: $[2 \\times 2]$\n",
-    "\n",
-    "We must also define *variable* tensors for the posterior - TensorFlow will allow these to change during the optimization in order to minimise the cost (free energy):\n",
-    "\n",
-    " - Posterior mean: $[2]$\n",
-    " - Posterior covariance: $[2 \\times 2]$\n",
-    "\n",
-    "The posterior covariance must be a positive-definite matrix - since the optimizer does not know this, it is possible that invalid values may arise and the optimization will fail. To get around this restriction we build the covariance matrix from its Chlolesky decomposition.\n",
-    "\n",
-    "$$\\textbf{C} = (\\textbf{C}_{chol})^T\\textbf{C}_{chol}$$\n",
-    "\n",
-    "$\\textbf{C}_{chol}$ must have positive diagonal elements, so we define the underlying variables for these elements in terms of the log and then form the full $\\textbf{C}_{chol}$ matrix as a sum of the exponentials of the log-diagonal elements, and independent variables for the off-diagonal components.\n",
-    "\n",
-    "The code for this is below. Note that we need to define initial values for the posterior variables. It turns out that in the stochastic method it is important that the initial poster variance is not too large so although the prior is not informative, the initial posterior is. \n",
-    "\n",
-    "The next requirement is to be able to obtain a sample of *predicted* data values from the posterior. In stochastic VB this is used to approximate the integrals in the calculation of the free energy. It is *not* related to the number of data samples we have. Smaller values give quicker calculation, but may result in a noisy, non-convergent optimization. We'll start off with a sample size of 5, but you can change this later if you want.\n",
-    "\n",
-    "Note the use of the 'reparameterization trick' to express the samples as the scaling of a fixed MVN distribution - this improves the ability of the optimizer to choose better values for the next iteration.\n",
-    "\n",
-    "Next we need to implement the free energy calculation. This is the sum of the reconstruction loss (the extent to which the posterior matches the data) and the latent loss (the extent to which the posterior matches the prior). Let's do the reconstruction loss first which is the expected value of the log-likelihood across the posterior distribution. Remember that the expectation integral is being approximated using the samples we have from the posterior. So we evaluate the log-likelihood for *each* set of samples and then take the mean across all the samples.\n",
-    "\n",
-    "On to the latent loss, this is the log-KL divergence between the posterior and prior. Since\n",
-    "both our prior and posterior are MVN distributions, we can use a known analytic result as\n",
-    "given in the tutorial:\n"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 4,
-   "metadata": {
-    "vscode": {
-     "languageId": "python"
-    }
-   },
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "2023-03-19 17:12:07.590629: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1510] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 38284 MB memory:  -> device: 0, name: NVIDIA A100-SXM4-40GB, pci bus id: 0000:03:00.0, compute capability: 8.0\n",
-      "2023-03-19 17:12:07.592452: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1510] Created device /job:localhost/replica:0/task:0/device:GPU:1 with 38284 MB memory:  -> device: 1, name: NVIDIA A100-SXM4-40GB, pci bus id: 0000:44:00.0, compute capability: 8.0\n",
-      "2023-03-19 17:12:07.594212: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1510] Created device /job:localhost/replica:0/task:0/device:GPU:2 with 38284 MB memory:  -> device: 2, name: NVIDIA A100-SXM4-40GB, pci bus id: 0000:84:00.0, compute capability: 8.0\n",
-      "2023-03-19 17:12:07.595828: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1510] Created device /job:localhost/replica:0/task:0/device:GPU:3 with 38284 MB memory:  -> device: 3, name: NVIDIA A100-SXM4-40GB, pci bus id: 0000:c4:00.0, compute capability: 8.0\n"
-     ]
-    }
-   ],
-   "source": [
-    "import tensorflow as tf\n",
-    "\n",
-    "data = tf.constant(DATA, dtype=tf.float32)\n",
-    "prior_means = tf.constant([m0, v0], dtype=tf.float32)\n",
-    "prior_covariance = tf.constant([[v0, 0.0], [0.0, w0]], dtype=tf.float32)\n",
-    "\n",
-    "chol_off_diag = tf.Variable([[0, 0], [0, 0]], dtype=tf.float32) ###variable\n",
-    "\n",
-    "post_means_init = [0.0, 0.0]\n",
-    "post_covariance_init = [[1.0, 0.0], [0.0, 1.0]]\n",
-    "\n",
-    "chol_log_diag = tf.Variable(tf.math.log(tf.linalg.diag_part(post_covariance_init)), dtype=tf.float32) ###variable\n",
-    "\n",
-    "post_means = tf.Variable(post_means_init, dtype=tf.float32)\n",
-    "\n",
-    "\n",
-    "\n",
-    "def cost_fun():\n",
-    "#   print(\"here\\n\")\n",
-    "    N=100\n",
-    "    S=5\n",
-    "# Comment in this line if you do NOT want to infer parameter covariances\n",
-    "#chol_off_diag = tf.constant([[0, 0], [0, 0]], dtype=tf.float32)\n",
-    "\n",
-    "    chol_diag = tf.linalg.diag(tf.math.sqrt(tf.math.exp(chol_log_diag)))\n",
-    "    post_covariance_chol = tf.math.add(chol_diag, tf.linalg.band_part(chol_off_diag, -1, 0))\n",
-    "\n",
-    "    post_covariance = tf.matmul(tf.transpose(post_covariance_chol), post_covariance_chol) \n",
-    "\n",
-    "    eps = tf.random.normal((2, S), 0, 1, dtype=tf.float32)\n",
-    "#   print(\"here\\n\")\n",
-    "# Start off each sample with the current posterior mean\n",
-    "# post_samples is now a tensor of shape [2, n_samples]\n",
-    "    samples = tf.tile(tf.reshape(post_means, [2, 1]), [1, S])\n",
-    "\n",
-    "# Now add the random sample scaled by the covariance\n",
-    "    post_samples = tf.add(samples, tf.matmul(post_covariance_chol, eps)) \n",
-    "\n",
-    "    mu_samples = post_samples[0]\n",
-    " #   print(\"here\\n\")\n",
-    "# Get the current estimate of the noise variance remembering that\n",
-    "# we are inferring the log of the noise precision, beta\n",
-    "    log_noise_var = -post_samples[1]\n",
-    "    noise_var = tf.math.exp(log_noise_var)\n",
-    "#    print(\"here\\n\")\n",
-    "# Each sample value predicts the full set of values in the data sample.\n",
-    "# For our constant-signal model, the prediction is simply a set of \n",
-    "# constant values. The prediction tensor will have shape [S, N]\n",
-    "# where S is the sample size and N is the number of data values\n",
-    "    prediction = tf.tile(tf.reshape(mu_samples, [S, 1]), [1, N])\n",
-    "#    print(\"here\\n\")\n",
-    "# To calculate the likelihood we need the sum of the squared difference between the data  \n",
-    "# and the prediction. This gives a value for each posterior sample so has shape [S]\n",
-    "#    print(data.shape, prediction.shape)\n",
-    "    sum_square_diff = tf.reduce_sum(tf.math.square(data - prediction), axis=-1)\n",
-    "    \n",
-    "# Now we calculate the likelihood for each posterior sample (shape [S])\n",
-    "# Note that we are ignoring constant factors such as 2*PI here as they \n",
-    "# are just an fixed offset and do not affect the optimization \n",
-    "    log_likelihood = 0.5 * (-log_noise_var * tf.cast(N,dtype=tf.float32) - sum_square_diff / noise_var)\n",
-    "\n",
-    "# Finally to evaluate the expectation value we take the mean across all the posterior\n",
-    "# samples\n",
-    "    reconstr_loss = -tf.reduce_mean(log_likelihood)\n",
-    "\n",
-    "    C = post_covariance\n",
-    "    C0 = prior_covariance\n",
-    "\n",
-    "    m_minus_m0 = tf.reshape(tf.subtract(post_means, prior_means), [-1, 1])\n",
-    "    m_minus_m0_T = tf.reshape(tf.subtract(post_means, prior_means), [1, -1])\n",
-    "\n",
-    "    C0_inv = tf.linalg.inv(C0)\n",
-    "\n",
-    "    term1 = tf.linalg.trace(tf.matmul(C0_inv, C))\n",
-    "    term2 = -tf.math.log(tf.linalg.det(C) / tf.linalg.det(C0))\n",
-    "\n",
-    "# Size of the MVN distribution\n",
-    "    term3 = -2\n",
-    "    term4 = tf.matmul(tf.matmul(m_minus_m0_T, C0_inv), m_minus_m0)\n",
-    "#   print(\"here\\n\")         \n",
-    "    latent_loss = 0.5 * (term1 + term2 + term3 + term4)\n",
-    " #   print(\"here\\n\")\n",
-    "    noise_var = tf.math.exp(log_noise_var)\n",
-    "\n",
-    "# Each sample value predicts the full set of values in the data sample.\n",
-    "# For our constant-signal model, the prediction is simply a set of \n",
-    "# constant values. The prediction tensor will have shape [S, N]\n",
-    "# where S is the sample size and N is the number of data values\n",
-    "#    print(\"here\\n\")\n",
-    "    prediction = tf.tile(tf.reshape(mu_samples, [S, 1]), [1, N])\n",
-    "#    print(prediction.shape,\" prediction shape\\n\")\n",
-    "\n",
-    "# To calculate the likelihood we need the sum of the squared difference between the data  \n",
-    "# and the prediction. This gives a value for each posterior sample so has shape [S]\n",
-    "    sum_square_diff = tf.reduce_sum(tf.math.square(data - prediction), axis=-1)\n",
-    "\n",
-    "# Now we calculate the likelihood for each posterior sample (shape [S])\n",
-    "# Note that we are ignoring constant factors such as 2*PI here as they \n",
-    "# are just an fixed offset and do not affect the optimization \n",
-    "    log_likelihood = 0.5 * (-log_noise_var * tf.cast(N,dtype=tf.float32) - sum_square_diff / noise_var)\n",
-    "\n",
-    "# Finally to evaluate the expectation value we take the mean across all the posterior\n",
-    "# samples\n",
-    "    reconstr_loss = -tf.reduce_mean(log_likelihood)\n",
-    "\n",
-    "    cost=reconstr_loss + latent_loss\n",
-    "\n",
-    "    return cost\n",
-    "\n",
-    "\n",
-    "    \n",
-    "    \n",
-    "\n"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "To optimize the posterior we need to iteratively minimise the cost using TensorFlow's built in gradient optimizer. We use the Adam optimizer for this, which is a refinement of the standard Gradient Descent optimizer. "
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 5,
-   "metadata": {
-    "vscode": {
-     "languageId": "python"
-    }
-   },
-   "outputs": [
-    {
-     "data": {
-      "application/vnd.jupyter.widget-view+json": {
-       "model_id": "4cac2bb6b7b14c3296e09193bc5b5d6e",
-       "version_major": 2,
-       "version_minor": 0
-      },
-      "text/plain": [
-       "Epoch::   0%|          | 0/5000 [00:00<?, ?it/s]"
-      ]
-     },
-     "metadata": {},
-     "output_type": "display_data"
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "2023-03-19 17:12:08.668453: I tensorflow/stream_executor/cuda/cuda_blas.cc:1760] TensorFloat-32 will be used for the matrix multiplication. This will only be logged once.\n",
-      "2023-03-19 17:12:08.672920: I tensorflow/core/util/cuda_solvers.cc:180] Creating CudaSolver handles for stream 0x2c9031d0\n"
-     ]
-    }
-   ],
-   "source": [
-    "optimizer = tf.keras.optimizers.Adam(learning_rate=0.5)\n",
-    "#minimizer = optimizer.minimize(cost)\n",
-    "#sess.run(tf.global_variables_initializer())\n",
-    "from tqdm.notebook import tqdm_notebook\n",
-    "\n",
-    "cost_history = []\n",
-    "for epoch in tqdm_notebook(range(5000), desc=\"Epoch:\"):\n",
-    "    #sess.run(minimizer)\n",
-    "\n",
-    "    #cost_history.append(float(sess.run(cost)))\n",
-    "    optimizer.minimize(cost_fun,var_list=[chol_off_diag,chol_log_diag,post_means])\n",
-    "    #print(float(cost_fun()))\n",
-    "    cost_history.append(float(cost_fun()))\n",
-    "    #print(\"Epoch %i: posterior means=%s\" % (epoch+1, post_means))"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "We should find our estimates for mu and beta are as expected, remembering that we chose\n",
-    "to estimate $-\\log{\\beta}$:"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 6,
-   "metadata": {
-    "vscode": {
-     "languageId": "python"
-    }
-   },
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Estimate for mu: 41.934135 (variance: 0.023306)\n",
-      "Estimate for beta: 0.962277\n"
-     ]
-    }
-   ],
-   "source": [
-    "final_means = post_means\n",
-    "chol_diag = tf.linalg.diag(tf.math.sqrt(tf.math.exp(chol_log_diag)))\n",
-    "post_covariance_chol = tf.math.add(chol_diag, tf.linalg.band_part(chol_off_diag, -1, 0))\n",
-    "post_covariance = tf.matmul(tf.transpose(post_covariance_chol), post_covariance_chol) \n",
-    "final_covariance = post_covariance\n",
-    "print(\"Estimate for mu: %f (variance: %f)\" % (final_means[0], final_covariance[0, 0]))\n",
-    "print(\"Estimate for beta: %f\" % np.exp(final_means[1]))"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "The convergence of the minimisation in this case is rather slow - considering the simplicity of the problem. This is partly due to the large difference between the initial posterior and the true value for $\\mu$. However, frameworks like TensorFlow are not really optimized for inferring such a small number of parameters and the method. If we plot the cost history we can see that it gets stuck in a local minimum for some time:"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 7,
-   "metadata": {
-    "vscode": {
-     "languageId": "python"
-    }
-   },
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "[<matplotlib.lines.Line2D at 0x1455f818c5e0>]"
-      ]
-     },
-     "execution_count": 7,
-     "metadata": {},
-     "output_type": "execute_result"
-    },
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 432x288 with 1 Axes>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "plt.plot(cost_history)\n",
-    "#plt.ylim(50000, 51000)"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "We can also plot our inferred value of $\\mu$ with error bars $\\pm 2\\sigma$ derived from the inferred variance on this parameter"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 8,
-   "metadata": {
-    "vscode": {
-     "languageId": "python"
-    }
-   },
-   "outputs": [
-    {
-     "data": {
-      "text/plain": [
-       "[<matplotlib.lines.Line2D at 0x1455f8125460>]"
-      ]
-     },
-     "execution_count": 8,
-     "metadata": {},
-     "output_type": "execute_result"
-    },
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 432x288 with 1 Axes>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "plt.figure()\n",
-    "plt.plot(DATA, \"rx\")\n",
-    "plt.plot([MU_TRUTH] * N, \"g\")\n",
-    "plt.plot([final_means[0]] * N, \"b\")\n",
-    "plt.plot([final_means[0]+2*np.sqrt(final_covariance[0, 0])] * N, \"b--\")\n",
-    "plt.plot([final_means[0]-2*np.sqrt(final_covariance[0, 0])] * N, \"b--\")"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "The main advantage of the SVB approach is the flexibility - as well as losing the restriction to conjugate priors, it is also much easier to implement more advanced types of parameters and priors, for example global parameters or spatial regularization priors. While these can (and have been) incorporated into the analytic VB framework, they require update equations to be re-derived whereas the SVB method simply needs an expression for the cost which is generally more straightforward.\n",
-    "\n",
-    "Some things you might like to try with this example:\n",
-    "\n",
-    " - Do not infer the covariance between the parameters (see commented out code in the definition of the posterior). This generally makes the convergence less noisy.\n",
-    " - Modify the number of samples and the learning rate and see how they affect the convergence\n",
-    " - Try implementing the stochastic form of the latent loss rather than the analytic result for an MVN that we have used here (see Box 1 in the tutorial). This is necessary in cases where we do not assume an MVN structure for the prior and posterior. Essentially you need to calculate the log PDF for the prior and posterior for each sample and take the mean over all the samples.\n",
-    " \n"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": []
-  }
- ],
- "metadata": {
-  "kernelspec": {
-   "display_name": "PyDeepLearning-1.1",
-   "language": "python",
-   "name": "pydeeplearning"
-  },
-  "language_info": {
-   "codemirror_mode": {
-    "name": "ipython",
-    "version": 3
-   },
-   "file_extension": ".py",
-   "mimetype": "text/x-python",
-   "name": "python",
-   "nbconvert_exporter": "python",
-   "pygments_lexer": "ipython3",
-   "version": "3.9.6"
-  },
-  "vscode": {
-   "interpreter": {
-    "hash": "31f2aee4e71d21fbe5cf8b01ff0e069b9275f58929596ceb00d14d90e3e16cd6"
-   }
-  }
- },
- "nbformat": 4,
- "nbformat_minor": 4
-}
diff --git a/BLcourse4/.ipynb_checkpoints/vae_mod-checkpoint.ipynb b/BLcourse4/.ipynb_checkpoints/vae_mod-checkpoint.ipynb
deleted file mode 100644
index 5e972f26142ec0a7fa2d8f512777350269eaf42c..0000000000000000000000000000000000000000
--- a/BLcourse4/.ipynb_checkpoints/vae_mod-checkpoint.ipynb
+++ /dev/null
@@ -1,618 +0,0 @@
-{
- "cells": [
-  {
-   "cell_type": "code",
-   "execution_count": 2,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "import time\n",
-    "import tensorflow as tf\n",
-    "from tensorflow.keras import layers\n",
-    "from IPython import display\n",
-    "import matplotlib.pyplot as plt\n",
-    "import numpy as np\n",
-    "%matplotlib inline\n",
-    "from tensorflow import keras\n",
-    "from keras import backend as K"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "## Load the MNIST dataset\n",
-    "Each MNIST image is originally a vector of 784 integers, each of which is between 0-255 and represents the intensity of a pixel. Model each pixel with a Bernoulli distribution in our model, and statically binarize the dataset.\n",
-    "\n",
-    "Use tf.data to batch and shuffle the data"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 3,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "2023-03-19 17:40:56.021384: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1613] Created device /job:localhost/replica:0/task:0/device:GPU:0 with 38278 MB memory:  -> device: 0, name: NVIDIA A100-SXM4-40GB, pci bus id: 0000:03:00.0, compute capability: 8.0\n",
-      "2023-03-19 17:40:56.023169: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1613] Created device /job:localhost/replica:0/task:0/device:GPU:1 with 38278 MB memory:  -> device: 1, name: NVIDIA A100-SXM4-40GB, pci bus id: 0000:44:00.0, compute capability: 8.0\n",
-      "2023-03-19 17:40:56.024878: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1613] Created device /job:localhost/replica:0/task:0/device:GPU:2 with 38278 MB memory:  -> device: 2, name: NVIDIA A100-SXM4-40GB, pci bus id: 0000:84:00.0, compute capability: 8.0\n",
-      "2023-03-19 17:40:56.026497: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1613] Created device /job:localhost/replica:0/task:0/device:GPU:3 with 38278 MB memory:  -> device: 3, name: NVIDIA A100-SXM4-40GB, pci bus id: 0000:c4:00.0, compute capability: 8.0\n"
-     ]
-    }
-   ],
-   "source": [
-    "(train_images, _), (test_images, _) = tf.keras.datasets.mnist.load_data(path='/p/project/training2305/course-material/mnist.npz')\n",
-    "def preprocess_images(images):\n",
-    "    images = images.reshape((images.shape[0], 28, 28, 1)) / 255.\n",
-    "    return np.where(images > .5, 1.0, 0.0).astype('float32')\n",
-    "\n",
-    "train_images = preprocess_images(train_images)\n",
-    "test_images = preprocess_images(test_images)\n",
-    "\n",
-    "train_size = 60000\n",
-    "batch_size = 32\n",
-    "test_size = 10000\n",
-    "\n",
-    "train_dataset = (tf.data.Dataset.from_tensor_slices(train_images)\n",
-    "                 .shuffle(train_size).batch(batch_size))\n",
-    "test_dataset = (tf.data.Dataset.from_tensor_slices(test_images)\n",
-    "                .shuffle(test_size).batch(batch_size))\n"
-   ]
-  },
-  {
-   "attachments": {
-    "image.png": {
-     "image/png": "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"
-    }
-   },
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "## Define the encoder and decoder networks with *tf.keras.Sequential*\n",
-    "\n",
-    "In this VAE example, use two s ConvNets for the encoder and decoder networks. In the literature, these networks are also referred to as inference/recognition and generative models respectively. Use `tf.keras.Sequential` to simplify implementation. Let $x$ and $z$ denote the observation and latent variable respectively in the following descriptions.\n",
-    "\n",
-    "### Encoder network\n",
-    "This defines the approximate posterior distribution $q(z|x)$, which takes as input an observation and outputs a set of parameters for specifying the conditional distribution of the latent representation $z$. \n",
-    "In this example, simply model the distribution as a diagonal Gaussian, and the network outputs the mean and log-variance parameters of a factorized Gaussian. \n",
-    "Output log-variance instead of the variance directly for numerical stability.\n",
-    "\n",
-    "### Decoder network \n",
-    "This defines the conditional distribution of the observation $p(x|z)$, which takes a latent sample $z$ as input and outputs the parameters for a conditional distribution of the observation.\n",
-    "Model the latent distribution prior $p(z)$ as a unit Gaussian.\n",
-    "\n",
-    "### Reparameterization trick\n",
-    "To generate a sample $z$ for the decoder during training, you can sample from the latent distribution defined by the parameters outputted by the encoder, given an input observation $x$.\n",
-    "However, this sampling operation creates a bottleneck because backpropagation cannot flow through a random node.\n",
-    "\n",
-    "To address this, use a reparameterization trick.\n",
-    "In our example, you approximate $z$ using the decoder parameters and another parameter $\\epsilon$ as follows:\n",
-    "\n",
-    "$$z = \\mu + \\sigma \\odot \\epsilon$$\n",
-    "\n",
-    "where $\\mu$ and $\\sigma$ represent the mean and standard deviation of a Gaussian distribution respectively. They can be derived from the decoder output. The $\\epsilon$ can be thought of as a random noise used to maintain stochasticity of $z$. Generate $\\epsilon$ from a standard normal distribution.\n",
-    "\n",
-    "The latent variable $z$ is now generated by a function of $\\mu$, $\\sigma$ and $\\epsilon$, which would enable the model to backpropagate gradients in the encoder through $\\mu$ and $\\sigma$ respectively, while maintaining stochasticity through $\\epsilon$.\n",
-    "\n",
-    "### Network architecture\n",
-    "In the encoder network there are a total of four Conv blocks each consisting of a `Conv2D`, `BatchNorm` and `LeakyReLU` activation function. In each block, the image is downsampled by a factor of two. The slope of `LeakyReLU` is by default 0.2.\n",
-    "\n",
-    "In the final block or the Flatten layer we convert the `[None, 7, 7, 64]` to a vector of size 3136.\n",
-    "\n",
-    "The decoder network of the variational autoencoder is exactly similar to a vanilla autoencoder. It takes an input of size `[None, 2]`. The initial block has a `Dense` layer having 3136 neurons, recall in the encoder function this was the size of the vector after flattening the output from the last conv block. There are a total of four Conv blocks. The Conv block `[1, 3]` consists of a `Conv2DTranspose`, `BatchNorm` and `LeakyReLU` activation function. The Conv block 4 has a `Conv2DTranspose` with sigmoid activation function, which squashes the output in the range `[0, 1]` since the images are normalized in that range. In each block, the image is upsampled by a factor of two.\n",
-    "\n",
-    "The output from the decoder network is a tensor of size `[None, 28, 28, 1]`.\n",
-    "![image.png](attachment:image.png)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 4,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "class CVAE(tf.keras.Model):\n",
-    "    def __init__(self, latent_dim):\n",
-    "        super(CVAE, self).__init__()\n",
-    "\n",
-    "        self.latent_dim=latent_dim\n",
-    "        self.encoder = tf.keras.Sequential(\n",
-    "    \n",
-    "    [\n",
-    "    tf.keras.layers.InputLayer(input_shape=(28,28,1), name='input_layer'),\n",
-    "    \n",
-    "    # Block-1\n",
-    "    tf.keras.layers.Conv2D(32, kernel_size=3, strides= 1, padding='same', name='conv_1'),\n",
-    "    tf.keras.layers.BatchNormalization(name='bn_1'),\n",
-    "    tf.keras.layers.LeakyReLU(name='lrelu_1'),\n",
-    "    \n",
-    "     \n",
-    "    # Block-2\n",
-    "    tf.keras.layers.Conv2D(64, kernel_size=3, strides= 2, padding='same', name='conv_2'),\n",
-    "    tf.keras.layers.BatchNormalization(name='bn_2'),\n",
-    "    tf.keras.layers.LeakyReLU(name='lrelu_2'),\n",
-    "     \n",
-    "    # Block-3\n",
-    "    tf.keras.layers.Conv2D(64, 3, 2, padding='same', name='conv_3'),\n",
-    "    tf.keras.layers.BatchNormalization(name='bn_3'),\n",
-    "    tf.keras.layers.LeakyReLU(name='lrelu_3'),\n",
-    "\n",
-    "   \n",
-    "    # Block-4\n",
-    "    tf.keras.layers.Conv2D(64, 3, 1, padding='same', name='conv_4'),\n",
-    "    tf.keras.layers.BatchNormalization(name='bn_4'),\n",
-    "    tf.keras.layers.LeakyReLU(name='lrelu_4'),    \n",
-    "\n",
-    "    # Final Block\n",
-    "    tf.keras.layers.Flatten(),\n",
-    "    tf.keras.layers.Dense(latent_dim+latent_dim, name='mean')\n",
-    "    ]\n",
-    "    )\n",
-    "\n",
-    "        self.decoder= tf.keras.Sequential(\n",
-    "          [\n",
-    "\n",
-    "    tf.keras.layers.InputLayer(input_shape=(latent_dim,), name='input_layer'),\n",
-    "    tf.keras.layers.Dense(3136, name='dense_1'),\n",
-    "    tf.keras.layers.Reshape((7, 7, 64), name='Reshape_Layer'),\n",
-    "    \n",
-    "    # Block-1\n",
-    "     tf.keras.layers.Conv2DTranspose(64, 3, strides= 1, padding='same',name='conv_transpose_1'),\n",
-    "     tf.keras.layers.BatchNormalization(name='bn_1'),\n",
-    "     tf.keras.layers.LeakyReLU(name='lrelu_1'),\n",
-    "   \n",
-    "    # Block-2\n",
-    "     tf.keras.layers.Conv2DTranspose(64, 3, strides= 2, padding='same', name='conv_transpose_2'),\n",
-    "     tf.keras.layers.BatchNormalization(name='bn_2'),\n",
-    "     tf.keras.layers.LeakyReLU(name='lrelu_2'),\n",
-    "     \n",
-    "    # Block-3\n",
-    "     tf.keras.layers.Conv2DTranspose(32, 3, 2, padding='same', name='conv_transpose_3'),\n",
-    "     tf.keras.layers.BatchNormalization(name='bn_3'),\n",
-    "     tf.keras.layers.LeakyReLU(name='lrelu_3'),\n",
-    "     \n",
-    "    # Block-4\n",
-    "     tf.keras.layers.Conv2DTranspose(1, 3, 1,padding='same', activation='sigmoid', name='conv_transpose_4')\n",
-    "          ]\n",
-    "     )\n",
-    "\n",
-    "    @tf.function\n",
-    "    def sample(self, eps=None):\n",
-    "        if eps is None:\n",
-    "            eps = tf.random.normal(shape=(100, self.latent_dim))\n",
-    "        return self.decode(eps, apply_sigmoid=True)\n",
-    "\n",
-    "    def encode(self, x):\n",
-    "        mean, logvar = tf.split(self.encoder(x), num_or_size_splits=2, axis=1)\n",
-    "        return mean, logvar\n",
-    "\n",
-    "    def reparameterize(self, mean, logvar):\n",
-    "        eps = tf.random.normal(shape=mean.shape)\n",
-    "        return eps * tf.exp(logvar * .5) + mean\n",
-    "    \n",
-    "    def decode(self, z, apply_sigmoid=False):\n",
-    "        logits = self.decoder(z)\n",
-    "        if apply_sigmoid:\n",
-    "            probs = tf.sigmoid(logits)\n",
-    "            return probs\n",
-    "        return logits\n",
-    "    \n",
-    "\n",
-    "\n"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "### Objective Function of VAE\n",
-    "\n",
-    "VAE’s loss function comprises a Reconstruction error, and a $KL$-divergence error used to model the networks’ objectives. The final loss is a weighted sum of both losses.\n",
-    "\n",
-    "VAE’s total loss can be given as:\n",
-    "\n",
-    "$L=(reconstruction\\ loss)+(regularization\\ term)$\n",
-    "\n",
-    "$L=\\frac1N\\sum_{i=1}^N(X_i-\\hat{X}_i)^2+KL(q(z|X)||N(0,1))$, where $X_i$s are input images, $\\hat{X}_i$ are the reconstructed images\n",
-    "\n",
-    "$KL(q(z|X)||N(0,1))=-\\frac12\\sum(1+\\log z^2_{\\sigma_i}-z^2_{\\mu_i}-z^2_{\\sigma_i})$\n",
-    "\n",
-    "### Training\n",
-    "\n",
-    "* Start by iterating over the dataset\n",
-    "* During each iteration, pass the image to the encoder to obtain a set of mean and log-variance parameters of the approximate posterior $q(z|x)$\n",
-    "* then apply the *reparameterization trick* to sample from $q(z|x)$\n",
-    "* Finally, pass the reparameterized samples to the decoder to obtain the logits of the generative distribution $p(x|z)$\n",
-    "* Note: Since you use the dataset loaded by keras with 60k datapoints in the training set and 10k datapoints in the test set, our resulting ELBO on the test set is slightly higher than reported results in the literature which uses dynamic binarization of Larochelle's MNIST.\n",
-    "\n",
-    "### Generating images\n",
-    "\n",
-    "* After training, it is time to generate some images\n",
-    "* Start by sampling a set of latent vectors from the unit Gaussian prior distribution $p(z)$\n",
-    "* The generator will then convert the latent sample $z$ to logits of the observation, giving a distribution $p(x|z)$\n"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 5,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "optimizer = tf.keras.optimizers.Adam(learning_rate = 0.0005)\n",
-    " \n",
-    "def mse_loss(y_true, y_pred):\n",
-    "    r_loss = K.mean(K.square(y_true - y_pred), axis = [1,2,3])\n",
-    "    return 1000 * r_loss\n",
-    " \n",
-    "def kl_loss(mean, log_var):\n",
-    "    kl_loss_val =  -0.5 * K.sum(1 + log_var - K.square(mean) - K.exp(log_var), axis = 1)\n",
-    "    return kl_loss_val\n",
-    " \n",
-    "def vae_loss(y_true, y_pred, mean, log_var):\n",
-    "    r_loss = mse_loss(y_true, y_pred)\n",
-    "    kl_loss_val = kl_loss(mean, log_var)\n",
-    "    #print(K.print_tensor(r_loss),\" loss\")\n",
-    "    return  r_loss + kl_loss_val\n",
-    "\n",
-    "def compute_loss(model,images):\n",
-    "        \n",
-    "    mean,log_var= model.encode(images)\n",
-    "    latent=model.reparameterize(mean,log_var)\n",
-    "    generated_images = model.decoder(latent)\n",
-    "    loss = vae_loss(images, generated_images, mean, log_var)\n",
-    "  \n",
-    "    return tf.reduce_mean(loss)\n",
-    "\n",
-    "\n",
-    "  \n",
-    "\n",
-    "def generate_and_save_images(model, epoch, test_sample):\n",
-    "    mean, logvar = model.encode(test_sample)\n",
-    "    z = model.reparameterize(mean, logvar)\n",
-    "    predictions = model.sample(z)\n",
-    "    fig = plt.figure(figsize=(4, 4))\n",
-    "\n",
-    "    for i in range(predictions.shape[0]):\n",
-    "        plt.subplot(4, 4, i + 1)\n",
-    "        plt.imshow(predictions[i, :, :, 0], cmap='gray')\n",
-    "        plt.axis('off')\n",
-    "\n",
-    "  # tight_layout minimizes the overlap between 2 sub-plots\n",
-    "    plt.savefig('image_at_epoch_{:04d}.png'.format(epoch))\n",
-    "    plt.show()"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "In the training loop we first pass the image to the encoder, then the latent variables mean and variance are fed to the sampling model and the output latent is finally fed to the decoder. The loss is computed over the images generated by the decoder.\n",
-    "\n",
-    "Next, we compute the gradients and update the encoder & decoder parameters using the Adam optimizer. Finally, we return the loss."
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 6,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "# Notice the use of `tf.function`\n",
-    "# This annotation causes the function to be \"compiled\".\n",
-    "@tf.function\n",
-    "def train_step(images,model,optimizer):\n",
-    "\n",
-    "    with tf.GradientTape() as tape:\n",
-    "        \n",
-    "        #mean,log_var= model.encode(images)\n",
-    "        #mean, log_var = tf.split(enc(images,training=True), num_or_size_splits=2, axis=1)\n",
-    "        #print(\" IS IT HERE???\")\n",
-    "        #latent = sampling(mean, log_var)\n",
-    "        #latent=model.reparameterize(mean,log_var)\n",
-    "        #generated_images = model.decoder(latent)\n",
-    "        #loss = vae_loss(images, generated_images, mean, log_var)\n",
-    "\n",
-    "        loss=compute_loss(model,images)\n",
-    " \n",
-    "    \n",
-    "\n",
-    "    gradients = tape.gradient(loss, model.trainable_variables)\n",
-    "     \n",
-    "     \n",
-    "    optimizer.apply_gradients(zip(gradients, model.trainable_variables))\n",
-    "    return loss"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "Finally, we train our Autoencoder model. The train function below takes the train_dataset, epochs,model, optimizer and test_sample as the parameters and calls the train_step function at every new batch a number of times.\n",
-    "\n",
-    "At every epoch it displays a generated image from the last training epoch using the test sample defined beforehand"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "2023-03-19 17:41:52.782596: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:428] Loaded cuDNN version 8600\n",
-      "2023-03-19 17:41:53.452299: I tensorflow/compiler/xla/stream_executor/cuda/cuda_blas.cc:630] TensorFloat-32 will be used for the matrix multiplication. This will only be logged once.\n"
-     ]
-    },
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 288x288 with 16 Axes>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "2023-03-19 17:41:56.945882: I tensorflow/compiler/xla/service/service.cc:173] XLA service 0x2ff81a40 initialized for platform CUDA (this does not guarantee that XLA will be used). Devices:\n",
-      "2023-03-19 17:41:56.945939: I tensorflow/compiler/xla/service/service.cc:181]   StreamExecutor device (0): NVIDIA A100-SXM4-40GB, Compute Capability 8.0\n",
-      "2023-03-19 17:41:56.945953: I tensorflow/compiler/xla/service/service.cc:181]   StreamExecutor device (1): NVIDIA A100-SXM4-40GB, Compute Capability 8.0\n",
-      "2023-03-19 17:41:56.945964: I tensorflow/compiler/xla/service/service.cc:181]   StreamExecutor device (2): NVIDIA A100-SXM4-40GB, Compute Capability 8.0\n",
-      "2023-03-19 17:41:56.945974: I tensorflow/compiler/xla/service/service.cc:181]   StreamExecutor device (3): NVIDIA A100-SXM4-40GB, Compute Capability 8.0\n",
-      "2023-03-19 17:41:56.950828: I tensorflow/compiler/mlir/tensorflow/utils/dump_mlir_util.cc:268] disabling MLIR crash reproducer, set env var `MLIR_CRASH_REPRODUCER_DIRECTORY` to enable.\n",
-      "2023-03-19 17:41:57.085920: I tensorflow/compiler/jit/xla_compilation_cache.cc:477] Compiled cluster using XLA!  This line is logged at most once for the lifetime of the process.\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "tf.Tensor(68.40858, shape=(), dtype=float32)  loss\n",
-      "Time for epoch 1 is 12.802140951156616 sec\n"
-     ]
-    },
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 288x288 with 16 Axes>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "tf.Tensor(70.2883, shape=(), dtype=float32)  loss\n",
-      "Time for epoch 2 is 7.563257694244385 sec\n"
-     ]
-    },
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 288x288 with 16 Axes>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "tf.Tensor(68.1349, shape=(), dtype=float32)  loss\n",
-      "Time for epoch 3 is 7.578803539276123 sec\n"
-     ]
-    },
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 288x288 with 16 Axes>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "tf.Tensor(71.223915, shape=(), dtype=float32)  loss\n",
-      "Time for epoch 4 is 7.584213495254517 sec\n"
-     ]
-    },
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 288x288 with 16 Axes>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "tf.Tensor(66.26627, shape=(), dtype=float32)  loss\n",
-      "Time for epoch 5 is 7.4921722412109375 sec\n"
-     ]
-    },
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 288x288 with 16 Axes>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "tf.Tensor(60.87711, shape=(), dtype=float32)  loss\n",
-      "Time for epoch 6 is 7.514315843582153 sec\n"
-     ]
-    },
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 288x288 with 16 Axes>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "tf.Tensor(66.79645, shape=(), dtype=float32)  loss\n",
-      "Time for epoch 7 is 7.503955364227295 sec\n"
-     ]
-    },
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 288x288 with 16 Axes>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "tf.Tensor(62.91003, shape=(), dtype=float32)  loss\n",
-      "Time for epoch 8 is 7.53310227394104 sec\n"
-     ]
-    },
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 288x288 with 16 Axes>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "tf.Tensor(61.488884, shape=(), dtype=float32)  loss\n",
-      "Time for epoch 9 is 7.558063745498657 sec\n"
-     ]
-    },
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 288x288 with 16 Axes>"
-      ]
-     },
-     "metadata": {
-      "needs_background": "light"
-     },
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "def train(dataset, epochs,model,optimizer,test_sample):\n",
-    "    generate_and_save_images(model, 0, test_sample)\n",
-    "    for epoch in range(epochs):\n",
-    "        start = time.time()\n",
-    "        for image_batch in dataset:\n",
-    "      \n",
-    "            l=train_step(image_batch,model,optimizer)\n",
-    "      \n",
-    "        print(l,\" loss\")\n",
-    "        print ('Time for epoch {} is {} sec'.format(epoch + 1, time.time()-start))\n",
-    "  \n",
-    "        generate_and_save_images(model, 0, test_sample)\n",
-    "\n",
-    "num_examples_to_generate = 16\n",
-    "\n",
-    "assert batch_size >= num_examples_to_generate\n",
-    "for test_batch in test_dataset.take(1):\n",
-    "    test_sample = test_batch[0:num_examples_to_generate, :, :, :]\n",
-    "\n",
-    "\n",
-    "\n",
-    "model=CVAE(2)\n",
-    "\n",
-    "\n",
-    "\n",
-    "train(train_dataset, 30,model,optimizer,test_sample)\n",
-    "\n",
-    "\n",
-    "generate_and_save_images(model, 0, test_sample)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": []
-  }
- ],
- "metadata": {
-  "kernelspec": {
-   "display_name": "sc_venv_template-bl",
-   "language": "python",
-   "name": "sc_venv_template-bl"
-  },
-  "language_info": {
-   "codemirror_mode": {
-    "name": "ipython",
-    "version": 3
-   },
-   "file_extension": ".py",
-   "mimetype": "text/x-python",
-   "name": "python",
-   "nbconvert_exporter": "python",
-   "pygments_lexer": "ipython3",
-   "version": "3.10.4"
-  }
- },
- "nbformat": 4,
- "nbformat_minor": 4
-}
diff --git a/BLcourse5/01_simple_gp_regression.ipynb b/BLcourse5/01_simple_gp_regression.ipynb
new file mode 100644
index 0000000000000000000000000000000000000000..b6581706745e41359849d48b271ce9febbd37d58
--- /dev/null
+++ b/BLcourse5/01_simple_gp_regression.ipynb
@@ -0,0 +1,496 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "id": "8bc26c14",
+   "metadata": {},
+   "source": [
+    "# Notation\n",
+    "$\\newcommand{\\ve}[1]{\\mathit{\\boldsymbol{#1}}}$\n",
+    "$\\newcommand{\\ma}[1]{\\mathbf{#1}}$\n",
+    "$\\newcommand{\\pred}[1]{\\widehat{#1}}$\n",
+    "$\\newcommand{\\cov}{\\mathrm{cov}}$\n",
+    "\n",
+    "Vector $\\ve a\\in\\mathbb R^n$ or $\\mathbb R^{n\\times 1}$, so \"column\" vector.\n",
+    "Matrix $\\ma A\\in\\mathbb R^{n\\times m}$. Design matrix with input vectors $\\ve\n",
+    "x_i\\in\\mathbb R^D$: $\\ma X = [\\ldots, \\ve x_i, \\ldots]^\\top \\in\\mathbb\n",
+    "R^{N\\times D}$.\n",
+    "\n",
+    "We use 1D data, so in fact $\\ma X \\in\\mathbb R^{N\\times 1}$ is a vector, but\n",
+    "we still denote the collection of all $\\ve x_i = x_i\\in\\mathbb R$ points with\n",
+    "$\\ma X$ to keep the notation consistent with the slides."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "3ef36e32",
+   "metadata": {},
+   "source": [
+    "# Imports, helpers, setup"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "c86d69b2",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "%matplotlib inline\n",
+    "\n",
+    "import math\n",
+    "from collections import defaultdict\n",
+    "from pprint import pprint\n",
+    "\n",
+    "import torch\n",
+    "import gpytorch\n",
+    "from matplotlib import pyplot as plt\n",
+    "from matplotlib import is_interactive\n",
+    "\n",
+    "\n",
+    "def extract_model_params(model, raw=False) -> dict:\n",
+    "    \"\"\"Helper to convert model.named_parameters() to dict.\n",
+    "\n",
+    "    With raw=True, use\n",
+    "        foo.bar.raw_param\n",
+    "    else\n",
+    "        foo.bar.param\n",
+    "\n",
+    "    See https://docs.gpytorch.ai/en/stable/examples/00_Basic_Usage/Hyperparameters.html#Raw-vs-Actual-Parameters\n",
+    "    \"\"\"\n",
+    "    if raw:\n",
+    "        return dict(\n",
+    "            (p_name, p_val.item())\n",
+    "            for p_name, p_val in model.named_parameters()\n",
+    "        )\n",
+    "    else:\n",
+    "        out = dict()\n",
+    "        # p_name = 'covar_module.base_kernel.raw_lengthscale'. Access\n",
+    "        # model.covar_module.base_kernel.lengthscale (w/o the raw_)\n",
+    "        for p_name, p_val in model.named_parameters():\n",
+    "            # Yes, eval() hack. Sorry.\n",
+    "            p_name = p_name.replace(\".raw_\", \".\")\n",
+    "            p_val = eval(f\"model.{p_name}\")\n",
+    "            out[p_name] = p_val.item()\n",
+    "        return out\n",
+    "\n",
+    "\n",
+    "def plot_samples(ax, X_pred, samples, label=None, **kwds):\n",
+    "    plot_kwds = dict(color=\"tab:green\", alpha=0.3)\n",
+    "    plot_kwds.update(kwds)\n",
+    "\n",
+    "    if label is None:\n",
+    "        ax.plot(X_pred, samples.T, **plot_kwds)\n",
+    "    else:\n",
+    "        ax.plot(X_pred, samples[0, :], **plot_kwds, label=label)\n",
+    "        ax.plot(X_pred, samples[1:, :].T, **plot_kwds, label=\"_\")\n",
+    "\n",
+    "\n",
+    "# Default float32 results in slightly noisy prior samples. Less so with\n",
+    "# float64. We get a warning with both\n",
+    "#   .../lib/python3.11/site-packages/linear_operator/utils/cholesky.py:40:\n",
+    "#       NumericalWarning: A not p.d., added jitter of 1.0e-08 to the diagonal\n",
+    "# but the noise is smaller w/ float64. Reason must be that the `sample()`\n",
+    "# method [1] calls `rsample()` [2] which performs a Cholesky decomposition of\n",
+    "# the covariance matrix. The default in\n",
+    "# np.random.default_rng().multivariate_normal() is method=\"svd\", which is\n",
+    "# slower but seemingly a bit more stable.\n",
+    "#\n",
+    "# [1] https://docs.gpytorch.ai/en/stable/distributions.html#gpytorch.distributions.MultivariateNormal.sample\n",
+    "# [2] https://docs.gpytorch.ai/en/stable/distributions.html#gpytorch.distributions.MultivariateNormal.rsample\n",
+    "torch.set_default_dtype(torch.float64)\n",
+    "\n",
+    "torch.manual_seed(123)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "9c543940",
+   "metadata": {
+    "lines_to_next_cell": 2
+   },
+   "source": [
+    "# Generate toy 1D data\n",
+    "\n",
+    "Here we generate noisy 1D data `X_train`, `y_train` as well as an extended\n",
+    "x-axis `X_pred` which we use later for prediction also outside of the data\n",
+    "range (extrapolation). The data has a constant offset `const` which we use to\n",
+    "test learning a GP mean function $m(\\ve x)$. We create a gap in the data to\n",
+    "show how the model uncertainty will behave there."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "0674cc5a",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def generate_data(x, gaps=[[1, 3]], const=5):\n",
+    "    y = torch.sin(x) * torch.exp(-0.2 * x) + torch.randn(x.shape) * 0.1 + const\n",
+    "    msk = torch.tensor([True] * len(x))\n",
+    "    if gaps is not None:\n",
+    "        for g in gaps:\n",
+    "            msk = msk & ~((x > g[0]) & (x < g[1]))\n",
+    "    return x[msk], y[msk]\n",
+    "\n",
+    "\n",
+    "x = torch.linspace(0, 4 * math.pi, 100)\n",
+    "X_train, y_train = generate_data(x, gaps=[[6, 10]])\n",
+    "X_pred = torch.linspace(\n",
+    "    X_train[0] - 2, X_train[-1] + 2, 200, requires_grad=False\n",
+    ")\n",
+    "\n",
+    "print(f\"{X_train.shape=}\")\n",
+    "print(f\"{y_train.shape=}\")\n",
+    "print(f\"{X_pred.shape=}\")\n",
+    "\n",
+    "plt.scatter(X_train, y_train, marker=\"o\", color=\"tab:blue\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "e4ddc32f",
+   "metadata": {
+    "lines_to_next_cell": 2
+   },
+   "source": [
+    "# Define GP model\n",
+    "\n",
+    "We define the simplest possible textbook GP model using a Gaussian\n",
+    "likelihood. The kernel is the squared exponential kernel with a scaling\n",
+    "factor.\n",
+    "\n",
+    "$$\\kappa(\\ve x_i, \\ve x_j) = \\sigma_f\\,\\exp\\left(-\\frac{\\lVert\\ve x_i - \\ve x_j\\rVert_2^2}{2\\,\\ell^2}\\right)$$\n",
+    "\n",
+    "This makes two hyper params, namely the length scale $\\ell$ and the scaling\n",
+    "$\\sigma_f$. The latter is implemented by wrapping the `RBFKernel` with\n",
+    "`ScaleKernel`.\n",
+    "\n",
+    "In addition, we define a constant mean via `ConstantMean`. Finally we have\n",
+    "the likelihood noise $\\sigma_n^2$. So in total we have 4 hyper params.\n",
+    "\n",
+    "* $\\ell$ = `model.covar_module.base_kernel.lengthscale`\n",
+    "* $\\sigma_n^2$ = `model.likelihood.noise_covar.noise`\n",
+    "* $\\sigma_f$ = `model.covar_module.outputscale`\n",
+    "* $m(\\ve x) = c$ = `model.mean_module.constant`"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "93822ff7",
+   "metadata": {
+    "lines_to_next_cell": 2
+   },
+   "outputs": [],
+   "source": [
+    "class ExactGPModel(gpytorch.models.ExactGP):\n",
+    "    \"\"\"API:\n",
+    "\n",
+    "    model.forward()             prior                   f_pred\n",
+    "    model()                     posterior               f_pred\n",
+    "\n",
+    "    likelihood(model.forward()) prior with noise        y_pred\n",
+    "    likelihood(model())         posterior with noise    y_pred\n",
+    "    \"\"\"\n",
+    "\n",
+    "    def __init__(self, X_train, y_train, likelihood):\n",
+    "        super().__init__(X_train, y_train, likelihood)\n",
+    "        self.mean_module = gpytorch.means.ConstantMean()\n",
+    "        self.covar_module = gpytorch.kernels.ScaleKernel(\n",
+    "            gpytorch.kernels.RBFKernel()\n",
+    "        )\n",
+    "\n",
+    "    def forward(self, x):\n",
+    "        mean_x = self.mean_module(x)\n",
+    "        covar_x = self.covar_module(x)\n",
+    "        return gpytorch.distributions.MultivariateNormal(mean_x, covar_x)\n",
+    "\n",
+    "\n",
+    "likelihood = gpytorch.likelihoods.GaussianLikelihood()\n",
+    "model = ExactGPModel(X_train, y_train, likelihood)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "3288a7ca",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Inspect the model\n",
+    "print(model)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "9348a1db",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Default start hyper params\n",
+    "pprint(extract_model_params(model, raw=False))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "b2b5d326",
+   "metadata": {
+    "lines_to_next_cell": 2
+   },
+   "outputs": [],
+   "source": [
+    "# Set new start hyper params\n",
+    "model.mean_module.constant = 3.0\n",
+    "model.covar_module.base_kernel.lengthscale = 1.0\n",
+    "model.covar_module.outputscale = 1.0\n",
+    "model.likelihood.noise_covar.noise = 0.1\n",
+    "\n",
+    "pprint(extract_model_params(model, raw=False))"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "c7309ee3",
+   "metadata": {},
+   "source": [
+    "# Sample from the GP prior\n",
+    "\n",
+    "We sample a number of functions $f_j, j=1,\\ldots,M$ from the GP prior and\n",
+    "evaluate them at all $\\ma X$ = `X_pred` points, of which we have $N'=200$. So\n",
+    "we effectively generate samples from $p(\\pred{\\ve y}|\\ma X) = \\mathcal N(\\ve\n",
+    "c, \\ma K)$. Each sampled vector $\\pred{\\ve y}\\in\\mathbb R^{N'}$ and the\n",
+    "covariance (kernel) matrix is $\\ma K\\in\\mathbb R^{N'\\times N'}$."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "56b26cd6",
+   "metadata": {
+    "lines_to_next_cell": 2
+   },
+   "outputs": [],
+   "source": [
+    "model.eval()\n",
+    "likelihood.eval()\n",
+    "\n",
+    "with torch.no_grad():\n",
+    "    # Prior\n",
+    "    M = 10\n",
+    "    pri_f = model.forward(X_pred)\n",
+    "    f_mean = pri_f.mean\n",
+    "    f_std = pri_f.stddev\n",
+    "    f_samples = pri_f.sample(sample_shape=torch.Size((M,)))\n",
+    "    print(f\"{pri_f=}\")\n",
+    "    print(f\"{pri_f.mean.shape=}\")\n",
+    "    print(f\"{pri_f.covariance_matrix.shape=}\")\n",
+    "    print(f\"{f_samples.shape=}\")\n",
+    "    fig, ax = plt.subplots()\n",
+    "    ax.plot(X_pred, f_mean, color=\"tab:red\", label=\"mean\", lw=2)\n",
+    "    plot_samples(ax, X_pred, f_samples, label=\"prior samples\")\n",
+    "    ax.fill_between(\n",
+    "        X_pred,\n",
+    "        f_mean - 2 * f_std,\n",
+    "        f_mean + 2 * f_std,\n",
+    "        color=\"tab:orange\",\n",
+    "        alpha=0.2,\n",
+    "        label=\"confidence\",\n",
+    "    )\n",
+    "    ax.legend()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "6758e47f",
+   "metadata": {},
+   "source": [
+    "Let's investigate the samples more closely. A constant mean $\\ve m(\\ma X) =\n",
+    "\\ve c$ does *not* mean that each sampled vector $\\pred{\\ve y}$'s mean is\n",
+    "equal to $c$. Instead, we have that at each $\\ve x_i$, the mean of\n",
+    "*all* sampled functions is the same, so $\\frac{1}{M}\\sum_{j=1}^M f_j(\\ve x_i)\n",
+    "\\approx c$ and for $M\\rightarrow\\infty$ it will be exactly $c$.\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "69ae529a",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Look at the first 20 x points from M=10 samples\n",
+    "print(f\"{f_samples.shape=}\")\n",
+    "print(f\"{f_samples.mean(axis=0)[:20]=}\")\n",
+    "print(f\"{f_samples.mean(axis=0).mean()=}\")\n",
+    "print(f\"{f_samples.mean(axis=0).std()=}\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "df46aa3f",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Take more samples, the means should get closer to c\n",
+    "f_samples = pri_f.sample(sample_shape=torch.Size((M * 200,)))\n",
+    "print(f\"{f_samples.shape=}\")\n",
+    "print(f\"{f_samples.mean(axis=0)[:20]=}\")\n",
+    "print(f\"{f_samples.mean(axis=0).mean()=}\")\n",
+    "print(f\"{f_samples.mean(axis=0).std()=}\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "5a9350d8",
+   "metadata": {},
+   "source": [
+    "# Fit GP to data: optimize hyper params\n",
+    "\n",
+    "In each step of the optimizer, we condition on the training data (e.g. do\n",
+    "Bayesian inference) to calculate the weight posterior for the current values\n",
+    "of the hyper params.\n",
+    "\n",
+    "We use a simplistic PyTorch-style hand written train loop without convergence\n",
+    "control, so make sure to use enough `n_iter` and eyeball-check that the loss\n",
+    "is converged :-)\n",
+    "\n",
+    "Observe how all hyper params converge. In particular, note that the constant\n",
+    "mean $m(\\ve x)=c$ converges to the `const` value in `generate_data()`."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "638d1244",
+   "metadata": {
+    "lines_to_next_cell": 2
+   },
+   "outputs": [],
+   "source": [
+    "# Train mode\n",
+    "model.train()\n",
+    "likelihood.train()\n",
+    "\n",
+    "optimizer = torch.optim.Adam(model.parameters(), lr=0.1)\n",
+    "loss_func = gpytorch.mlls.ExactMarginalLogLikelihood(likelihood, model)\n",
+    "\n",
+    "n_iter = 200\n",
+    "history = defaultdict(list)\n",
+    "for ii in range(n_iter):\n",
+    "    optimizer.zero_grad()\n",
+    "    loss = -loss_func(model(X_train), y_train)\n",
+    "    loss.backward()\n",
+    "    optimizer.step()\n",
+    "    if (ii + 1) % 10 == 0:\n",
+    "        print(f\"iter {ii+1}/{n_iter}, {loss=:.3f}\")\n",
+    "    for p_name, p_val in extract_model_params(model).items():\n",
+    "        history[p_name].append(p_val)\n",
+    "    history[\"loss\"].append(loss.item())\n",
+    "\n",
+    "# Plot hyper params and loss (neg. log marginal likelihood) convergence\n",
+    "ncols = len(history)\n",
+    "fig, axs = plt.subplots(ncols=ncols, nrows=1, figsize=(ncols * 5, 5))\n",
+    "for ax, (p_name, p_lst) in zip(axs, history.items()):\n",
+    "    ax.plot(p_lst)\n",
+    "    ax.set_title(p_name)\n",
+    "    ax.set_xlabel(\"iterations\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "777b67d9",
+   "metadata": {},
+   "source": [
+    "# Run prediction\n",
+    "\n",
+    "We show \"noiseless\" (left: $\\sigma = \\sqrt{\\mathrm{diag}(\\ma\\Sigma)}$) vs.\n",
+    "\"noisy\" (right: $\\sigma = \\sqrt{\\mathrm{diag}(\\ma\\Sigma + \\sigma_n^2\\,\\ma\n",
+    "I_N)}$) predictions, where $\\ma\\Sigma\\equiv\\cov(\\ve f_*)$ is the posterior\n",
+    "predictive covariance matrix from R&W 2006 eq. 2.24 with $\\ma K = K(X,X)$,\n",
+    "$\\ma K'=K(X_*, X)$ and $\\ma K''=K(X_*, X_*)$, so\n",
+    "\n",
+    "$$\\ma\\Sigma = \\ma K'' - \\ma K'\\,(\\ma K+\\sigma_n^2\\,\\ma I)^{-1}\\,\\ma K'^\\top$$\n",
+    "\n",
+    "See\n",
+    "https://elcorto.github.io/gp_playground/content/gp_pred_comp/notebook_plot.html\n",
+    "for details."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "740fb21a",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Evaluation (predictive posterior) mode\n",
+    "model.eval()\n",
+    "likelihood.eval()\n",
+    "\n",
+    "with torch.no_grad():\n",
+    "    post_pred_f = model(X_pred)\n",
+    "    post_pred_y = likelihood(model(X_pred))\n",
+    "\n",
+    "    fig, axs = plt.subplots(ncols=2, figsize=(12, 5))\n",
+    "    for ii, (ax, post_pred) in enumerate(zip(axs, [post_pred_f, post_pred_y])):\n",
+    "        yf_mean = post_pred.mean\n",
+    "        yf_samples = post_pred.sample(sample_shape=torch.Size((10,)))\n",
+    "\n",
+    "        ##lower, upper = post_pred.confidence_region()\n",
+    "        yf_std = post_pred.stddev\n",
+    "        lower = yf_mean - 2 * yf_std\n",
+    "        upper = yf_mean + 2 * yf_std\n",
+    "        ax.plot(\n",
+    "            X_train.numpy(),\n",
+    "            y_train.numpy(),\n",
+    "            \"o\",\n",
+    "            label=\"data\",\n",
+    "            color=\"tab:blue\",\n",
+    "        )\n",
+    "        ax.plot(\n",
+    "            X_pred.numpy(),\n",
+    "            yf_mean.numpy(),\n",
+    "            label=\"mean\",\n",
+    "            color=\"tab:red\",\n",
+    "            lw=2,\n",
+    "        )\n",
+    "        ax.fill_between(\n",
+    "            X_pred.numpy(),\n",
+    "            lower.numpy(),\n",
+    "            upper.numpy(),\n",
+    "            label=\"confidence\",\n",
+    "            color=\"tab:orange\",\n",
+    "            alpha=0.3,\n",
+    "        )\n",
+    "        y_min = y_train.min()\n",
+    "        y_max = y_train.max()\n",
+    "        y_span = y_max - y_min\n",
+    "        ax.set_ylim([y_min - 0.3 * y_span, y_max + 0.3 * y_span])\n",
+    "        plot_samples(ax, X_pred, yf_samples, label=\"posterior pred. samples\")\n",
+    "        if ii == 1:\n",
+    "            ax.legend()\n",
+    "\n",
+    "# When running as script\n",
+    "if not is_interactive():\n",
+    "    plt.show()"
+   ]
+  }
+ ],
+ "metadata": {
+  "jupytext": {
+   "formats": "ipynb,py:percent"
+  },
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "language": "python",
+   "name": "python3"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}
diff --git a/BLcourse5/notebook.py b/BLcourse5/01_simple_gp_regression.py
similarity index 99%
rename from BLcourse5/notebook.py
rename to BLcourse5/01_simple_gp_regression.py
index a04c73ceb63975409a70b54c429ea4762fc72312..fcfcda544ec40c9be131cde8bb0937b0980f886c 100644
--- a/BLcourse5/notebook.py
+++ b/BLcourse5/01_simple_gp_regression.py
@@ -13,7 +13,6 @@
 #     name: python3
 # ---
 
-
 # %% [markdown]
 # # Notation
 # $\newcommand{\ve}[1]{\mathit{\boldsymbol{#1}}}$
diff --git a/BLcourse5/README.md b/BLcourse5/README.md
index 1ff90a31fe8b20349f3b966999a22bf6091a8b09..21766d2c861fc568cce33e935241b908c3ce8938 100644
--- a/BLcourse5/README.md
+++ b/BLcourse5/README.md
@@ -6,3 +6,6 @@ Use https://jupytext.readthedocs.io to create a notebook.
 $ jupytext --to ipynb notebook.py
 $ jupyter-notebook notebook.ipynb
 ```
+
+For convenience the ipynb file is included. Please keep it in sync with the py
+file using jupytext.