diff --git a/mlair/plotting/postprocessing_plotting.py b/mlair/plotting/postprocessing_plotting.py
index 6cf92c08a8d3d81ac57eaa0e3618c8eb8ace67c7..77bf7aef01ffcf7bae6f19f8f57966b11bf8b321 100644
--- a/mlair/plotting/postprocessing_plotting.py
+++ b/mlair/plotting/postprocessing_plotting.py
@@ -18,6 +18,7 @@ import seaborn as sns
import xarray as xr
from matplotlib.backends.backend_pdf import PdfPages
from matplotlib.offsetbox import AnchoredText
+import matplotlib as mpl
from scipy.stats import mannwhitneyu
from mlair import helpers
@@ -1453,6 +1454,172 @@ class PlotSampleUncertaintyFromBootstrap(AbstractPlotClass): # pragma: no cover
return ax
+@TimeTrackingWrapper
+class PlotScoresOnMap(AbstractPlotClass):
+ from mlair.plotting.data_insight_plotting import PlotStationMap
+
+ def __init__(self, plot_folder: str, skill_score_report_path: str, model_name: str,
+ reference_models: Union[str, List[str]]):
+
+ super().__init__(plot_folder, f"map_plot_skill_scores")
+ skill_score_report_path = "/media/felix/INTENSO/2021-12-03_1008-08_WRF_sector3_network_daily/latex_report"
+ reference_models = ['ols', 'NNb', 'NN1s', 'persi' ]
+ self.skill_score_report_path = skill_score_report_path
+ self.model_name = model_name
+ self.reference_models = reference_models
+
+ ### internal data
+ self._data = None
+ self._ax = None
+ self._gl = None
+
+
+ self._plot()
+ ### workflow
+ # self.open_data()
+ # self._coords = self.get_coords_from_index()
+
+ def open_data(self, file_name):
+ # read csv data
+ df = pd.read_csv(file_name)
+ # drop "total" column
+ # df = df[df.index != "total"].copy()
+ df = df[df[df.columns[0]] != "total"].copy()
+ df[["lat", "lon"]] = df[df.columns[0]].str.split("__", expand=True).replace("_", ".").replace("_", ".",
+ regex=True).iloc[
+ :, 1:].astype(float)
+
+ df = df.set_index(df.columns[0])
+ # df = df.set_index(["lat", "lon"])
+ return df.astype(float)
+
+ def _plot_individual(self):
+ import cartopy.feature as cfeature
+ import cartopy.crs as ccrs
+ from cartopy.mpl.gridliner import LONGITUDE_FORMATTER, LATITUDE_FORMATTER
+ from mpl_toolkits.axes_grid1 import make_axes_locatable
+
+ for competitor in self.reference_models:
+ file_name = os.path.join(self.skill_score_report_path,
+ f"error_report_skill_score_{self.model_name}_-_{competitor}.csv"
+ )
+
+ plot_path = os.path.join(os.path.abspath(self.plot_folder),
+ f"{self.plot_name}_{self.model_name}_-_{competitor}.pdf")
+ pdf_pages = matplotlib.backends.backend_pdf.PdfPages(plot_path)
+ df = self.open_data(file_name)
+
+ bounds = np.linspace(-1, 1, 100)
+ cmap = mpl.cm.coolwarm
+ norm = mpl.colors.BoundaryNorm(bounds, cmap.N, extend='both')
+ ticks = np.arange(norm.vmin, norm.vmax + .2, .2)
+ for i, lead_name in enumerate(df.columns[:-2]): # last two are lat lon
+ fig = plt.figure()
+ self._ax = fig.add_subplot(1, 1, 1, projection=ccrs.PlateCarree())
+
+ self._ax.add_feature(cfeature.LAND.with_scale("50m"))
+ # self._ax.natural_earth_shp(resolution='50m')
+ self._ax.add_feature(cfeature.COASTLINE.with_scale("50m"), edgecolor='black')
+ # self._ax.add_feature(cfeature.LAKES.with_scale("50m"))
+ self._ax.add_feature(cfeature.OCEAN.with_scale("50m"))
+ self._ax.add_feature(cfeature.RIVERS.with_scale("50m"))
+ self._ax.add_feature(cfeature.BORDERS.with_scale("50m"), facecolor='none', edgecolor='black')
+ self._ax.scatter(df.lon.values, df.lat.values, c=df[lead_name],
+ transform=ccrs.PlateCarree(),
+ norm=norm, cmap=cmap)
+ self._gl = self._ax.gridlines(xlocs=range(0, 21, 5), ylocs=range(44, 59, 2), draw_labels=True)
+ self._gl.xformatter = LONGITUDE_FORMATTER
+ self._gl.yformatter = LATITUDE_FORMATTER
+ label = f"Skill Score: {lead_name.replace('-', 'vs.').replace('(t+', ' (').replace(')', 'd)')}"
+ # cax = fig.add_axes([self._ax.get_position().x0,
+ # self._ax.get_position().y0-.02,
+ # self._ax.get_position().width,
+ # 0.02])
+ # cax = fig.add_axes([0.09, 0.06, 0.84, 0.02])
+ self._cbar = fig.colorbar(mpl.cm.ScalarMappable(norm=norm, cmap=cmap),
+ orientation='horizontal', ticks=ticks,
+ label=label,
+ # cax=cax
+ )
+ # self._ax.set_aspect('auto')
+ # plt.tight_layout()
+ # pdf_pages.savefig()
+ # close all open figures / plots
+ pdf_pages.savefig()
+ pdf_pages.close()
+ plt.close('all')
+
+ def _plot(self, ncol: int = 2):
+ import cartopy.feature as cfeature
+ import cartopy.crs as ccrs
+ from cartopy.mpl.gridliner import LONGITUDE_FORMATTER, LATITUDE_FORMATTER
+ import string
+ base_plot_name = self.plot_name
+ for competitor in self.reference_models:
+ file_name = os.path.join(self.skill_score_report_path,
+ f"error_report_skill_score_{self.model_name}_-_{competitor}.csv"
+ )
+
+ self.plot_name = f"{base_plot_name}_{self.model_name}_-_{competitor}"
+ df = self.open_data(file_name)
+
+ nrow = int(np.ceil(len(df.columns[:-2])/ncol))
+ bounds = np.linspace(-1, 1, 100)
+ cmap = mpl.cm.coolwarm
+ norm = mpl.colors.BoundaryNorm(bounds, cmap.N, extend='both')
+ ticks = np.arange(norm.vmin, norm.vmax + .2, .2)
+ fig, self._axes = plt.subplots(nrows=nrow, ncols=ncol, subplot_kw={'projection': ccrs.PlateCarree()})
+ for i, ax in enumerate(self._axes.reshape(-1)): # last two are lat lon
+
+ sub_name = f"({string.ascii_lowercase[i]})"
+ lead_name = df.columns[i]
+ ax.add_feature(cfeature.LAND.with_scale("50m"))
+ ax.add_feature(cfeature.COASTLINE.with_scale("50m"), edgecolor='black')
+ ax.add_feature(cfeature.OCEAN.with_scale("50m"))
+ ax.add_feature(cfeature.RIVERS.with_scale("50m"))
+ ax.add_feature(cfeature.BORDERS.with_scale("50m"), facecolor='none', edgecolor='black')
+ ax.scatter(df.lon.values, df.lat.values, c=df[lead_name],
+ marker='.',
+ transform=ccrs.PlateCarree(),
+ norm=norm, cmap=cmap)
+ gl = ax.gridlines(xlocs=range(0, 21, 5), ylocs=range(44, 59, 2), draw_labels=True)
+ gl.xformatter = LONGITUDE_FORMATTER
+ gl.yformatter = LATITUDE_FORMATTER
+ gl.top_labels = []
+ gl.right_labels = []
+ ax.text(0.01, 1.09, f'{sub_name} {lead_name.split("+")[1][:-1]}d',
+ verticalalignment='top', horizontalalignment='left',
+ transform=ax.transAxes,
+ color='black',
+ )
+ label = f"Skill Score: {lead_name.replace('-', 'vs.').split('(')[0]}"
+
+ fig.subplots_adjust(bottom=0.18)
+ cax = fig.add_axes([0.15, 0.1, 0.7, 0.02])
+ self._cbar = fig.colorbar(mpl.cm.ScalarMappable(norm=norm, cmap=cmap),
+ orientation='horizontal',
+ ticks=ticks,
+ label=label,
+ cax=cax
+ )
+
+ fig.subplots_adjust(wspace=.001, hspace=.2)
+ self._save(bbox_inches="tight")
+ plt.close('all')
+
+
+ @staticmethod
+ def get_coords_from_index(name_string: str) -> List[float]:
+ """
+
+ :param name_string:
+ :type name_string:
+ :return: List of coords [lat, lon]
+ :rtype: List
+ """
+ res = [float(frac.replace("_", ".")) for frac in name_string.split(sep="__")[1:]]
+ return res
+
if __name__ == "__main__":
stations = ['DEBW107', 'DEBY081', 'DEBW013', 'DEBW076', 'DEBW087']
path = "../../testrun_network/forecasts"
diff --git a/mlair/run_modules/post_processing.py b/mlair/run_modules/post_processing.py
index 5985db43fcd35491122de9739de9f607c0e3909e..f71e55c8d2c5182d6c3e104758e0ffe42e148263 100644
--- a/mlair/run_modules/post_processing.py
+++ b/mlair/run_modules/post_processing.py
@@ -24,7 +24,7 @@ from mlair.model_modules.linear_model import OrdinaryLeastSquaredModel
from mlair.model_modules import AbstractModelClass
from mlair.plotting.postprocessing_plotting import PlotMonthlySummary, PlotClimatologicalSkillScore, \
PlotCompetitiveSkillScore, PlotTimeSeries, PlotFeatureImportanceSkillScore, PlotConditionalQuantiles, \
- PlotSeparationOfScales, PlotSampleUncertaintyFromBootstrap, PlotSectorialSkillScore
+ PlotSeparationOfScales, PlotSampleUncertaintyFromBootstrap, PlotSectorialSkillScore, PlotScoresOnMap
from mlair.plotting.data_insight_plotting import PlotStationMap, PlotAvailability, PlotAvailabilityHistogram, \
PlotPeriodogram, PlotDataHistogram
from mlair.run_modules.run_environment import RunEnvironment
@@ -691,6 +691,16 @@ class PostProcessing(RunEnvironment):
logging.error(f"Could not create plot PlotSampleUncertaintyFromBootstrap due to the following error: {e}"
f"\n{sys.exc_info()[0]}\n{sys.exc_info()[1]}\n{sys.exc_info()[2]}")
+ try:
+ if "PlotScoresOnMap" in plot_list:
+ report_path = os.path.join(self.data_store.get("experiment_path"), "latex_report")
+ PlotScoresOnMap(plot_folder=self.plot_path, skill_score_report_path=report_path,
+ model_name=self.model_display_name,
+ reference_models=self.competitors+["persi, ols"])
+ except Exception as e:
+ logging.error(f"Could not create plot PlotScoresOnMap due to the following error: {e}"
+ f"\n{sys.exc_info()[0]}\n{sys.exc_info()[1]}\n{sys.exc_info()[2]}")
+
def calculate_test_score(self):
"""Evaluate test score of model and save locally."""