[refactor] Ruff fixes

docs/source/conf.py

@@ -15,13 +15,14 @@
 import os
 import sys
 
-import sphinx_fontawesome
+import sphinx_fontawesome  # noqa: F401
 from sphinx.ext.autodoc import between
 
+from typing import Any, Dict
+
 # sys.path.insert(0, os.path.abspath('.'))
 sys.path.insert(0, os.path.abspath("../.."))
 
-from typing import Any, Dict
 
 # -- Project information -----------------------------------------------------
 
@@ -71,12 +72,14 @@
 html_favicon = "images/np_favicon.png"
 html_logo = "images/np_highres_docs.svg"
 # html_logo = "images/logo.png"
+font_stack = "-apple-system,'system-ui','Segoe UI',Helvetica,Arial,sans-serif,'Apple Color Emoji','Segoe UI Emoji'"
+font_stack_mono = "'SFMono-Regular',Menlo,Consolas,Monaco,Liberation Mono,Lucida Console,monospace"
 html_theme_options: Dict[str, Any] = {
     "sidebar_hide_name": True,
     "navigation_with_keys": True,
     "light_css_variables": {
-        "font-stack": "-apple-system, 'system-ui', 'Segoe UI', Helvetica, Arial, sans-serif, 'Apple Color Emoji', 'Segoe UI Emoji'",
-        "font-stack--monospace": "'SFMono-Regular',Menlo,Consolas,Monaco,Liberation Mono,Lucida Console,monospace",
+        "font-stack": font_stack,
+        "font-stack--monospace": font_stack_mono,
     },
 }
 

neuralprophet/components/future_regressors/linear.py

@@ -37,7 +37,8 @@ def scalar_features_effects(self, features, params, indices=None):
         Parameters
         ----------
             features : torch.Tensor, float
-                Features (either additive or multiplicative) related to event component dims (batch, n_forecasts, n_features)
+                Features (either additive or multiplicative) related to event component dims (batch, n_forecasts,
+                n_features)
             params : nn.Parameter
                 Params (either additive or multiplicative) related to events
             indices : list of int

neuralprophet/components/seasonality/fourier.py

@@ -23,7 +23,8 @@ def __init__(self, config, id_list, quantiles, num_seasonalities_modelled, n_for
             # Seasonality parameters for global or local modelling
             self.season_params = nn.ParameterDict(
                 {
-                    # dimensions - [no. of quantiles, num_seasonalities_modelled, no. of fourier terms for each seasonality]
+                    # dimensions -
+                    # [no. of quantiles, num_seasonalities_modelled, no. of fourier terms for each seasonality]
                     name: init_parameter(dims=[len(self.quantiles)] + [self.num_seasonalities_modelled] + [dim])
                     for name, dim in self.season_dims.items()
                 }

neuralprophet/components/trend/piecewise_linear.py

@@ -111,9 +111,11 @@ def add_regularization(self):
         pass
 
     def compute_k_t(self, current_segment, past_next_changepoint, meta_name_tensor_one_hot):
-        """For segmentwise, k_t is the model parameter representing the trend slope(actually, trend slope-k_0) in the current_segment at time t (for each sample of the batch).
+        """For segmentwise, k_t is the model parameter representing the trend slope(actually, trend slope-k_0) in the
+        current_segment at time t (for each sample of the batch).
 
-        For not segmentwise, k_t is the model parameter representing the difference between trend slope in the current_segment at time t and the trend slope in the previous segment (for each sample of the batch).
+        For not segmentwise, k_t is the model parameter representing the difference between trend slope in the
+        current_segment at time t and the trend slope in the previous segment (for each sample of the batch).
 
         Parameters
         ----------
@@ -137,12 +139,9 @@ def compute_k_t(self, current_segment, past_next_changepoint, meta_name_tensor_o
         pass
 
     def compute_m_t(self, current_segment, past_next_changepoint, meta_name_tensor_one_hot):
-        """m_t represents the value at the origin(t=0) that we would need to have so that if we use (k_t + k_0) as slope,
-
-        we reach the same value at time = chagepoint_start_of_segment_i
-
-        as we would reach by following the segmented slope (having in each segment the slope trend_deltas(i) + k_0)
-
+        """m_t represents the value at the origin(t=0) that we would need to have so that if we use (k_t + k_0) as
+        slope, we reach the same value at time = chagepoint_start_of_segment_i as we would reach by following the
+        segmented slope (having in each segment the slope trend_deltas(i) + k_0)
 
         Parameters
         ----------
@@ -206,16 +205,17 @@ def __init__(self, config, id_list, quantiles, num_trends_modelled, n_forecasts,
 
     def compute_k_t(self, current_segment, past_next_changepoint, meta_name_tensor_one_hot=None):
         """This method overrides the method from the PiecewiseLinear class."""
-        # For segmentwise k_t, is the model parameter representing the trend slope(actually, trend slope-k_0) in the current_segment at time t (for each sample of the batch).
+        # For segmentwise k_t, is the model parameter representing the trend slope(actually, trend slope-k_0) in the
+        # current_segment at time t (for each sample of the batch).
         # k_t = k_t(current_segment).
         # dimensions - batch_size, n_forecasts, quantiles_size
         k_t = torch.sum(
             current_segment.unsqueeze(dim=2) * self.trend_deltas.permute(1, 0, 2).unsqueeze(1),
             dim=-1,
         )
 
-        # For not segmentwise k_t is the model parameter representing the difference between trend slope in the current_segment at time t
-        # and the trend slope in the previous segment (for each sample of the batch).
+        # For not segmentwise k_t is the model parameter representing the difference between trend slope in the
+        # current_segment at time t and the trend slope in the previous segment (for each sample of the batch).
         if not self.segmentwise_trend:
             # k_t = k_t(current_segment, previous_segment)
             # dimensions - batch_size, n_forecasts, quantiles_size
@@ -252,7 +252,8 @@ def compute_m_t(self, current_segment, past_next_changepoint, meta_name_tensor_o
             if not self.segmentwise_trend:
                 m_t = m_t.detach()
         else:
-            # For discontinuous, trend_m is a parameter to optimize, as it is not defined just by trend_deltas & trend_k0
+            # For discontinuous, trend_m is a parameter to optimize, as it is not defined just
+            # by trend_deltas & trend_k0
             # m_t = m_t(current_segment)
             # dimensions - batch_size, n_forecasts, quantiles
             m_t = torch.sum(current_segment.unsqueeze(dim=2) * self.trend_m.permute(1, 0, 2).unsqueeze(dim=0), dim=-1)
@@ -279,7 +280,8 @@ def __init__(self, config, id_list, quantiles, num_trends_modelled, n_forecasts,
 
     def compute_k_t(self, current_segment, past_next_changepoint, meta_name_tensor_one_hot):
         """This method overrides the method from the PiecewiseLinear class."""
-        # For segmentwise k_t, is the model parameter representing the trend slope(actually, trend slope-k_0) in the current_segment at time t (for each sample of the batch).
+        # For segmentwise k_t, is the model parameter representing the trend slope(actually, trend slope-k_0) in the
+        # current_segment at time t (for each sample of the batch).
         # k_t = k_t(current_segment, sample metadata)
         # dimensions - quantiles, batch_size, segments (+ 1)
         trend_deltas_by_sample = torch.sum(
@@ -288,8 +290,8 @@ def compute_k_t(self, current_segment, past_next_changepoint, meta_name_tensor_o
         # dimensions - batch_size, n_forecasts, quantiles_size
         k_t = torch.sum(current_segment.unsqueeze(dim=2) * trend_deltas_by_sample.permute(1, 0, 2).unsqueeze(1), dim=-1)
 
-        # For not segmentwise k_t is the model parameter representing the difference between trend slope in the current_segment at time t
-        # and the trend slope in the previous segment (for each sample of the batch).
+        # For not segmentwise k_t is the model parameter representing the difference between trend slope in the
+        # current_segment at time t and the trend slope in the previous segment (for each sample of the batch).
         if not self.segmentwise_trend:
             # k_t = k_t(current_segment, previous_segment, sample metadata)
             previous_deltas_t = torch.sum(
@@ -335,7 +337,8 @@ def compute_m_t(self, current_segment, past_next_changepoint, meta_name_tensor_o
             if not self.segmentwise_trend:
                 m_t = m_t.detach()
         else:
-            # For discontinuous, trend_m is a parameter to optimize, as it is not defined just by trend_deltas & trend_k0
+            # For discontinuous, trend_m is a parameter to optimize, as it is not defined just
+            # by trend_deltas & trend_k0
             # m_t = m_t(current_segment, sample metadata)
             # dimensions - quantiles, batch_size, segments
             m_t_0 = torch.sum(

neuralprophet/configure.py

@@ -65,12 +65,14 @@ def get_data_params(self, df_name):
                 data_params = self.local_data_params[df_name]
             elif self.unknown_data_normalization:
                 log.debug(
-                    f"Dataset name {df_name!r} is not present in valid data_params but unknown_data_normalization is True. Using global_data_params"
+                    f"Dataset name {df_name!r} is not present in valid data_params but unknown_data_normalization is \
+                        True. Using global_data_params"
                 )
                 data_params = self.global_data_params
             else:
                 raise ValueError(
-                    f"Dataset name {df_name!r} missing from training data params. Set unknown_data_normalization to use global (average) normalization parameters."
+                    f"Dataset name {df_name!r} missing from training data params. Set unknown_data_normalization to \
+                        use global (average) normalization parameters."
                 )
         return data_params
 
@@ -177,8 +179,8 @@ def set_auto_batch_epoch(
 
     def set_optimizer(self):
         """
-        Set the optimizer and optimizer args. If optimizer is a string, then it will be converted to the corresponding torch optimizer.
-        The optimizer is not initialized yet as this is done in configure_optimizers in TimeNet.
+        Set the optimizer and optimizer args. If optimizer is a string, then it will be converted to the corresponding
+        torch optimizer. The optimizer is not initialized yet as this is done in configure_optimizers in TimeNet.
         """
         self.optimizer, self.optimizer_args = utils_torch.create_optimizer_from_config(
             self.optimizer, self.optimizer_args
@@ -207,7 +209,8 @@ def set_lr_finder_args(self, dataset_size, num_batches):
         num_training = 150 + int(np.log10(100 + dataset_size) * 25)
         if num_batches < num_training:
             log.warning(
-                f"Learning rate finder: The number of batches ({num_batches}) is too small than the required number for the learning rate finder ({num_training}). The results might not be optimal."
+                f"Learning rate finder: The number of batches ({num_batches}) is too small than the required number \
+                    for the learning rate finder ({num_training}). The results might not be optimal."
             )
             # num_training = num_batches
         self.lr_finder_args.update(

neuralprophet/data/process.py

@@ -451,7 +451,8 @@ def _handle_missing_data(
 
             Note
             ----
-            Any valid frequency for pd.date_range, such as ``5min``, ``D``, ``MS`` or ``auto`` (default) to automatically set frequency.
+            Any valid frequency for pd.date_range, such as ``5min``, ``D``, ``MS`` or ``auto`` (default) to
+            automatically set frequency.
     n_lags : int
         Previous time series steps to include in auto-regression. Aka AR-order
     n_forecasts : int
@@ -521,7 +522,8 @@ def _handle_missing_data_single_id(
 
             Note
             ----
-            Any valid frequency for pd.date_range, such as ``5min``, ``D``, ``MS`` or ``auto`` (default) to automatically set frequency.
+            Any valid frequency for pd.date_range, such as ``5min``, ``D``, ``MS`` or ``auto`` (default) to
+            automatically set frequency.
     n_lags : int
         Previous time series steps to include in auto-regression. Aka AR-order
     n_forecasts : int
@@ -559,10 +561,13 @@ def _handle_missing_data_single_id(
         if missing_dates > 0:
             if config_missing.impute_missing:
                 log.info(f"{missing_dates} missing dates added.")
-            # FIX Issue#52
-            # Comment error raising to allow missing data for autoregression flow.
-            # else:
-            #     raise ValueError(f"{missing_dates} missing dates found. Please preprocess data manually or set impute_missing to True.")
+                # FIX Issue#52
+                # Comment error raising to allow missing data for autoregression flow.
+                # else:
+                #     raise ValueError(
+                #         f"{missing_dates} missing dates found. Please preprocess data manually or set \
+                #             impute_missing to True."
+                #     )
             # END FIX
 
     if config_regressors is not None:
@@ -646,7 +651,8 @@ def _handle_missing_data_single_id(
                 log.info(f"{sum_na - remaining_na} NaN values in column {column} were auto-imputed.")
                 if remaining_na > 0:
                     log.warning(
-                        f"More than {2 * config_missing.impute_linear + config_missing.impute_rolling} consecutive missing values encountered in column {column}. "
+                        f"More than {2 * config_missing.impute_linear + config_missing.impute_rolling} consecutive \
+                            missing values encountered in column {column}. "
                         f"{remaining_na} NA remain after auto-imputation. "
                     )
             # FIX Issue#52

neuralprophet/data/split.py

@@ -227,13 +227,13 @@ def _make_future_dataframe(
         if nan_at_end > 0:
             if max_lags > 0 and (nan_at_end + 1) >= max_lags:
                 raise ValueError(
-                    f"{nan_at_end + 1} missing values were detected at the end of df before df was extended into the future. "
-                    "Please make sure there are no NaN values at the end of df."
+                    f"{nan_at_end + 1} missing values were detected at the end of df before df was extended into "
+                    "the future. Please make sure there are no NaN values at the end of df."
                 )
             df["y"].iloc[-(nan_at_end + 1) :].ffill(inplace=True)
             log.warning(
-                f"{nan_at_end + 1} missing values were forward-filled at the end of df before df was extended into the future. "
-                "Please make sure there are no NaN values at the end of df."
+                f"{nan_at_end + 1} missing values were forward-filled at the end of df before df was extended into the "
+                "future. Please make sure there are no NaN values at the end of df."
             )
 
     if len(df) > 0: