diff --git a/tests/test_input_checks.py b/tests/test_input_checks.py
index eb1691aa5135aa985cd12c494902bd8a5e3a9d96..9cad7f733160a3940a5536895d30ae27ca36ffc3 100644
--- a/tests/test_input_checks.py
+++ b/tests/test_input_checks.py
@@ -4,12 +4,25 @@ import numpy as np
import pandas as pd
import pytest
-from toarstats.input_checks import (
- check_crops, check_data, check_data_capture, check_index,
- check_input_parameters, check_metadata, check_required_parameters,
- check_sampling, check_seasons, check_station_climatic_zone,
- check_station_latitude, check_station_longitude, check_statistics,
- check_values, from_pandas, is_correct_type, is_empty, is_in_range,
+from toarstats.metrics.input_checks import (
+ check_crops,
+ check_data,
+ check_data_capture,
+ check_index,
+ check_input_parameters,
+ check_metadata,
+ check_required_parameters,
+ check_sampling,
+ check_seasons,
+ check_station_climatic_zone,
+ check_station_latitude,
+ check_station_longitude,
+ check_statistics,
+ check_values,
+ from_pandas,
+ is_correct_type,
+ is_empty,
+ is_in_range,
to_collection
)
@@ -260,16 +273,14 @@ class TestCheckData:
start="2000", periods=3, freq="H", tz="Europe/Berlin"
)
values = np.array([5., 6.1, 7])
- expected = pd.DataFrame(
- {"values": values}, index=index.tz_localize(None)
- )
- pd.testing.assert_frame_equal(
+ expected = pd.Series(values, index=index.tz_localize(None))
+ pd.testing.assert_series_equal(
check_data(pd.DataFrame(values, index=index), None, None), expected
)
- pd.testing.assert_frame_equal(
+ pd.testing.assert_series_equal(
check_data(pd.Series(values, index=index), None, None), expected
)
- pd.testing.assert_frame_equal(
+ pd.testing.assert_series_equal(
check_data(None, index, values), expected
)
@@ -593,9 +604,8 @@ class TestCheckInputParameters:
assert error_msg == str(excinfo.value)
def test_check_input_parameters_correct_input(self):
- data = pd.DataFrame(
- {"values": range(10)},
- index=pd.date_range(start="2000", periods=10, freq="H")
+ data = pd.Series(
+ range(10), index=pd.date_range(start="2000", periods=10, freq="H")
)
expected_required = namedtuple(
"Required",
@@ -604,7 +614,7 @@ class TestCheckInputParameters:
result = check_input_parameters("annual", "median", data, *[None]*9)
assert result.sampling == "annual"
assert result.statistics == ["median"]
- pd.testing.assert_frame_equal(result.data, data)
+ pd.testing.assert_series_equal(result.data, data)
assert result.metadata is None
assert result.seasons is None
assert result.crops is None
diff --git a/tests/test_interface.py b/tests/test_interface.py
index d476de3d0246ae820e959c799e6efcd41fd58657..45c1d1ac920e89ed70e209d1ce9bbdff003b50db 100644
--- a/tests/test_interface.py
+++ b/tests/test_interface.py
@@ -1,7 +1,7 @@
import numpy as np
import pandas as pd
-from toarstats.interface import calculate_statistics
+from toarstats.metrics.interface import calculate_statistics
class TestCalculateStatistics:
diff --git a/tests/test_stats.py b/tests/test_stats.py
index 90caa08e1543205bdaa25b6a7dd18b2494a0625e..a1d169d20425a94dfbbb516877d81633ba40b35c 100644
--- a/tests/test_stats.py
+++ b/tests/test_stats.py
@@ -2,16 +2,20 @@ import numpy as np
import pandas as pd
import pytest
-from toarstats.constants import ALLOWED_SAMPLING_VALUES, RSTAGS, SEASON_DICT
-from toarstats.defaults import DEFAULT_DATA_CAPTURE
-from toarstats.interface import calculate_statistics
-from toarstats.stats_utils import create_reference_data_frame
+from toarstats.metrics.constants import (
+ ALLOWED_SAMPLING_VALUES,
+ RSTAGS,
+ SEASON_DICT
+)
+from toarstats.metrics.defaults import DEFAULT_DATA_CAPTURE
+from toarstats.metrics.interface import calculate_statistics
+from toarstats.metrics.stats_utils import create_reference_series
data = pd.read_csv(
"tests/time_series.csv", header=None, names=[None, "values"],
index_col=0, parse_dates=True, infer_datetime_format=True
)
-ref_data = create_reference_data_frame(data.index)
+ref_data = create_reference_series(data.index)
metadata = {"station_lat": 50.906389,
"station_lon": 6.403889,
"station_climatic_zone": "cool temperate moist"}
@@ -52,12 +56,12 @@ def test_data_capture():
cur_ref.index += pd.Timedelta(182, "days")
count = cur_data.resample(offset).count().squeeze("columns")
expected[name] = count.divide(
- cur_ref.resample(offset).count().squeeze("columns")
+ cur_ref.resample(offset).count()
).reindex(count.index)
else:
count = data.resample(offset).count().squeeze("columns")
expected["data_capture"] = count.divide(
- ref_data.resample(offset).count().squeeze("columns")
+ ref_data.resample(offset).count()
).reindex(count.index)
pd.testing.assert_frame_equal(result, expected)
@@ -132,7 +136,7 @@ def test_median():
cur_ref.index += pd.Timedelta(182, "days")
data_rs = cur_data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = cur_ref.resample(offset).count().squeeze("columns")
+ ref_count = cur_ref.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.median().reindex(frac.index)
tmp_res[frac < DEFAULT_DATA_CAPTURE] = np.nan
@@ -140,7 +144,7 @@ def test_median():
else:
data_rs = data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = ref_data.resample(offset).count().squeeze("columns")
+ ref_count = ref_data.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.median().reindex(frac.index)
tmp_res[frac < DEFAULT_DATA_CAPTURE] = np.nan
@@ -164,7 +168,7 @@ def test_maximum():
cur_ref.index += pd.Timedelta(182, "days")
data_rs = cur_data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = cur_ref.resample(offset).count().squeeze("columns")
+ ref_count = cur_ref.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.max().reindex(frac.index)
tmp_res[frac < DEFAULT_DATA_CAPTURE] = np.nan
@@ -172,7 +176,7 @@ def test_maximum():
else:
data_rs = data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = ref_data.resample(offset).count().squeeze("columns")
+ ref_count = ref_data.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.max().reindex(frac.index)
tmp_res[frac < DEFAULT_DATA_CAPTURE] = np.nan
@@ -196,7 +200,7 @@ def test_minimum():
cur_ref.index += pd.Timedelta(182, "days")
data_rs = cur_data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = cur_ref.resample(offset).count().squeeze("columns")
+ ref_count = cur_ref.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.min().reindex(frac.index)
tmp_res[frac < DEFAULT_DATA_CAPTURE] = np.nan
@@ -204,7 +208,7 @@ def test_minimum():
else:
data_rs = data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = ref_data.resample(offset).count().squeeze("columns")
+ ref_count = ref_data.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.min().reindex(frac.index)
tmp_res[frac < DEFAULT_DATA_CAPTURE] = np.nan
@@ -228,7 +232,7 @@ def test_perc05():
cur_ref.index += pd.Timedelta(182, "days")
data_rs = cur_data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = cur_ref.resample(offset).count().squeeze("columns")
+ ref_count = cur_ref.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.quantile(0.05).reindex(frac.index)
tmp_res[frac < DEFAULT_DATA_CAPTURE] = np.nan
@@ -236,7 +240,7 @@ def test_perc05():
else:
data_rs = data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = ref_data.resample(offset).count().squeeze("columns")
+ ref_count = ref_data.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.quantile(0.05).reindex(frac.index)
tmp_res[frac < DEFAULT_DATA_CAPTURE] = np.nan
@@ -260,7 +264,7 @@ def test_perc10():
cur_ref.index += pd.Timedelta(182, "days")
data_rs = cur_data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = cur_ref.resample(offset).count().squeeze("columns")
+ ref_count = cur_ref.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.quantile(0.1).reindex(frac.index)
tmp_res[frac < DEFAULT_DATA_CAPTURE] = np.nan
@@ -268,7 +272,7 @@ def test_perc10():
else:
data_rs = data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = ref_data.resample(offset).count().squeeze("columns")
+ ref_count = ref_data.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.quantile(0.1).reindex(frac.index)
tmp_res[frac < DEFAULT_DATA_CAPTURE] = np.nan
@@ -292,7 +296,7 @@ def test_perc25():
cur_ref.index += pd.Timedelta(182, "days")
data_rs = cur_data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = cur_ref.resample(offset).count().squeeze("columns")
+ ref_count = cur_ref.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.quantile(0.25).reindex(frac.index)
tmp_res[frac < DEFAULT_DATA_CAPTURE] = np.nan
@@ -300,7 +304,7 @@ def test_perc25():
else:
data_rs = data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = ref_data.resample(offset).count().squeeze("columns")
+ ref_count = ref_data.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.quantile(0.25).reindex(frac.index)
tmp_res[frac < DEFAULT_DATA_CAPTURE] = np.nan
@@ -324,7 +328,7 @@ def test_perc75():
cur_ref.index += pd.Timedelta(182, "days")
data_rs = cur_data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = cur_ref.resample(offset).count().squeeze("columns")
+ ref_count = cur_ref.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.quantile(0.75).reindex(frac.index)
tmp_res[frac < DEFAULT_DATA_CAPTURE] = np.nan
@@ -332,7 +336,7 @@ def test_perc75():
else:
data_rs = data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = ref_data.resample(offset).count().squeeze("columns")
+ ref_count = ref_data.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.quantile(0.75).reindex(frac.index)
tmp_res[frac < DEFAULT_DATA_CAPTURE] = np.nan
@@ -356,7 +360,7 @@ def test_perc90():
cur_ref.index += pd.Timedelta(182, "days")
data_rs = cur_data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = cur_ref.resample(offset).count().squeeze("columns")
+ ref_count = cur_ref.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.quantile(0.9).reindex(frac.index)
tmp_res[frac < DEFAULT_DATA_CAPTURE] = np.nan
@@ -364,7 +368,7 @@ def test_perc90():
else:
data_rs = data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = ref_data.resample(offset).count().squeeze("columns")
+ ref_count = ref_data.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.quantile(0.9).reindex(frac.index)
tmp_res[frac < DEFAULT_DATA_CAPTURE] = np.nan
@@ -388,7 +392,7 @@ def test_perc95():
cur_ref.index += pd.Timedelta(182, "days")
data_rs = cur_data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = cur_ref.resample(offset).count().squeeze("columns")
+ ref_count = cur_ref.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.quantile(0.95).reindex(frac.index)
tmp_res[frac < DEFAULT_DATA_CAPTURE] = np.nan
@@ -396,7 +400,7 @@ def test_perc95():
else:
data_rs = data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = ref_data.resample(offset).count().squeeze("columns")
+ ref_count = ref_data.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.quantile(0.95).reindex(frac.index)
tmp_res[frac < DEFAULT_DATA_CAPTURE] = np.nan
@@ -420,7 +424,7 @@ def test_perc98():
cur_ref.index += pd.Timedelta(182, "days")
data_rs = cur_data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = cur_ref.resample(offset).count().squeeze("columns")
+ ref_count = cur_ref.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.quantile(0.98).reindex(frac.index)
tmp_res[frac < DEFAULT_DATA_CAPTURE] = np.nan
@@ -428,7 +432,7 @@ def test_perc98():
else:
data_rs = data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = ref_data.resample(offset).count().squeeze("columns")
+ ref_count = ref_data.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.quantile(0.98).reindex(frac.index)
tmp_res[frac < DEFAULT_DATA_CAPTURE] = np.nan
@@ -452,7 +456,7 @@ def test_perc99():
cur_ref.index += pd.Timedelta(182, "days")
data_rs = cur_data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = cur_ref.resample(offset).count().squeeze("columns")
+ ref_count = cur_ref.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.quantile(0.99).reindex(frac.index)
tmp_res[frac < DEFAULT_DATA_CAPTURE] = np.nan
@@ -460,7 +464,7 @@ def test_perc99():
else:
data_rs = data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = ref_data.resample(offset).count().squeeze("columns")
+ ref_count = ref_data.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.quantile(0.99).reindex(frac.index)
tmp_res[frac < DEFAULT_DATA_CAPTURE] = np.nan
@@ -484,7 +488,7 @@ def test_percentiles1():
cur_ref.index += pd.Timedelta(182, "days")
data_rs = cur_data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = cur_ref.resample(offset).count().squeeze("columns")
+ ref_count = cur_ref.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.quantile(
int(name.split("-", 1)[0][1:]) / 100
@@ -494,7 +498,7 @@ def test_percentiles1():
else:
data_rs = data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = ref_data.resample(offset).count().squeeze("columns")
+ ref_count = ref_data.resample(offset).count()
frac = data_count / ref_count
for name in result.columns.to_list():
tmp_res = data_rs.quantile(int(name[1:]) / 100).reindex(
@@ -521,7 +525,7 @@ def test_percentiles2():
cur_ref.index += pd.Timedelta(182, "days")
data_rs = cur_data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = cur_ref.resample(offset).count().squeeze("columns")
+ ref_count = cur_ref.resample(offset).count()
frac = data_count / ref_count
tmp_res = data_rs.quantile(
int(name.split("-", 1)[0][1:]) / 100
@@ -531,7 +535,7 @@ def test_percentiles2():
else:
data_rs = data.resample(offset)
data_count = data_rs.count().squeeze("columns")
- ref_count = ref_data.resample(offset).count().squeeze("columns")
+ ref_count = ref_data.resample(offset).count()
frac = data_count / ref_count
for name in result.columns.to_list():
tmp_res = data_rs.quantile(int(name[1:]) / 100).reindex(
diff --git a/tests/test_toarstats.py b/tests/test_toarstats.py
index ce994d04c79331d8fb8c0799b8d9b39bce87a85c..609346700db3d5dcd512c1f7417909795d6a0d9e 100644
--- a/tests/test_toarstats.py
+++ b/tests/test_toarstats.py
@@ -23,7 +23,8 @@ import numpy as np
import pandas as pd
import pytest
-from toarstats.interface import calculate_statistics
+from tests.create_sample_data_and_reference_results import create_sample_data
+from toarstats.metrics.interface import calculate_statistics
def get_all_statistics():
@@ -33,11 +34,11 @@ def get_all_statistics():
"""
statistics = set()
for file in Path(Path(__file__).resolve().parents[1],
- "toarstats").glob("*.py"):
+ "toarstats/metrics").glob("*.py"):
for node in ast.parse(file.read_text(), file).body:
if (isinstance(node, ast.FunctionDef)
and [el.arg for el in node.args.args]
- == ["df", "dfref", "mtype", "metadata", "seasons",
+ == ["ser", "ref", "mtype", "metadata", "seasons",
"data_capture"]):
statistics.add(node.name)
return statistics
@@ -50,7 +51,7 @@ def get_all_samplings():
"""
samplings = set()
for file in Path(Path(__file__).resolve().parents[1],
- "toarstats").glob("*.py"):
+ "toarstats/metrics").glob("*.py"):
for node in ast.parse(file.read_text(), file).body:
if (isinstance(node, ast.Assign)
and isinstance(node.value, ast.Dict)
@@ -68,8 +69,13 @@ def sample_data():
:return: A data frame with sample data
"""
+ sample_data_file = Path(
+ Path(__file__).resolve().parent, "sample_data/sample_data.csv"
+ )
+ if not sample_data_file.is_file():
+ create_sample_data(sample_data_file.parent)
return pd.read_csv(
- Path(Path(__file__).resolve().parent, "sample_data/sample_data.csv"),
+ sample_data_file,
header=None, index_col=0, parse_dates=True
)