calculate symmetry using Mira et al

optim_esm_tools/analyze/time_statistics.py

@@ -106,17 +106,26 @@ def calculate_skewtest(ds, field=None):
     return scipy.stats.skewtest(values, nan_policy='omit').pvalue
 
 
-def calculate_historical_std(ds, field='std detrended', **kw):
+def calculate_symmetry_test(ds, field=None):
+    import rpy_symmetry as rsym
+
+    values = get_values_from_data_set(ds, field)
+    return rsym.p_symmetry(values)
+
+
+def calculate_historical_std(
+    ds, field='std detrended', field_pi_control='max jump', **kw
+):
     ds_hist = get_historical_ds(ds, **kw)
     if ds_hist is None:
         return None
     mask = get_mask_from_global_mask(ds)
     ds_hist_masked = oet.analyze.xarray_tools.mask_xr_ds(ds_hist, mask, drop=True)
-    assert (
-        ds[field].shape == ds_hist_masked[field].shape
-    ), f'{ds[field].shape} != {ds_hist_masked[field].shape}'
-    cur = ds[field].values
-    his = ds_hist_masked[field].values
+    da = ds[field]
+    da_hist = ds_hist_masked[field_pi_control]
+    assert da.shape == da_hist.shape, f'{da.shape} != {da_hist.shape}'
+    cur = da.values
+    his = da_hist.values
     isnnan = np.isnan(cur) | np.isnan(his)
     cur = cur[~isnnan]
     his = his[~isnnan]

requirements.txt

@@ -7,3 +7,5 @@ pandas
 tqdm
 treelib
 xarray
+# Optional, required for some submodules:
+# rpy_symmetry

requirements_tests.txt

@@ -9,5 +9,6 @@ hypothesis-numpy
 intake-esm
 numba
 pytest
+rpy_symmetry
 scikit-learn
 scipy