linting

fedeca/utils/moments_utils.py

@@ -26,7 +26,6 @@ def compute_uncentered_moment(data, order):
     NotImplementedError
         Raised if the data type is not Dataframe nor np.ndarray.
     """
-
     if isinstance(data, (pd.DataFrame, pd.Series)):
         moment = data.select_dtypes(include=np.number).pow(order).mean(skipna=True)
     elif isinstance(data, np.ndarray):
@@ -99,7 +98,6 @@ def aggregation_mean(local_means: List[Any], n_local_samples: List[int]):
     Any
         Aggregated mean. Same type of the local means
     """
-
     tot_samples = np.nan_to_num(np.copy(n_local_samples[0]), nan=0, copy=False)
     tot_mean = np.nan_to_num(np.copy(local_means[0]), nan=0, copy=False)
     for mean, n_sample in zip(local_means[1:], n_local_samples[1:]):

fedeca/utils/survival_utils.py

@@ -1588,8 +1588,9 @@ def build_X_y_function(
     shared_state={},
     global_moments={},
 ):
-    """Build the inputs for a propensity model and for a Cox model as well as y directly
-    from the output of the opener.
+    """Build the inputs for a propensity model and for a Cox model and y.
+
+    Does that directly on data from opener.
 
     This function 1. uses the event column to inject the censorship
     information present in the duration column (given in absolute values)
@@ -1729,7 +1730,7 @@ def compute_X_y_and_propensity_weights_function(
     if propensity_model is not None:
         assert (
             treated is not None
-        ), f"""If you are using a propensity model the Treated
+        ), """If you are using a propensity model the Treated
         column should be available"""
         assert np.all(
             np.in1d(np.unique(treated.astype("uint8"))[0], [0, 1])