Hide decimator internals (Avaiga#1959)

* Added a sample for decimation
* Change chart mode to markers for demoing decimator zoom facility
* Hide decimator internals for the time begin
* Add return types where needed

doc/gui/examples/charts/advanced_large_datasets.py

@@ -79,7 +79,7 @@ class Processing(Enum):
 # Create the DataFrame
 data = pd.DataFrame({"X": x_values, "Y": y_values_noise})
 
-page = "<|{data}|chart|x=X|y=Y|mode=lines|decimator=decimator|>"
+page = "<|{data}|chart|x=X|y=Y|mode=markers|decimator=decimator|>"
 
 if __name__ == "__main__":
     Gui(page).run(title="Chart - Advanced - Large datasets")

taipy/gui/data/decimator/base.py

@@ -138,7 +138,7 @@ def _df_apply_decimator(
             x_column_name = self.__get_indexed_df_col(df)
         column_list = [x_column_name, y_column_name, z_column_name] if z_column_name else [x_column_name, y_column_name]
         points = df[column_list].to_numpy()
-        mask = self.decimate(points, payload)
+        mask = self._decimate(points, payload)
         return df[mask], is_copied
 
     def _on_decimate_df(
@@ -171,7 +171,7 @@ def _on_decimate_df(
         is_decimator_applied = False
         if nb_rows_max and self._is_applicable(df, nb_rows_max, chart_mode):
             try:
-                df, is_copied = self.apply_decimator(
+                df, is_copied = self._apply_decimator(
                     t.cast(pd.DataFrame, df),
                     x_column,
                     y_column,
@@ -188,16 +188,15 @@ def _on_decimate_df(
             df = df.filter(filterd_columns, axis=1)
         return df, is_decimator_applied, is_copied
 
-    def on_decimate(
+    def _on_decimate(
         self,
         df: pd.DataFrame,
         decimator_instance_payload: t.Dict[str, t.Any],
         decimator_payload: t.Dict[str, t.Any],
         is_copied: bool = False,
         filter_unused_columns: bool = True,
-    ):
+    ) -> t.Tuple:
         """NOT DOCUMENTED
-
         This function is executed whenever a decimator is found during runtime.
 
         Users can override this function by providing an alternate implementation inside the constructor
@@ -225,15 +224,15 @@ def on_decimate(
                 _warn("Error executing user defined on_decimate function: ", e)
         return self._on_decimate_df(df, decimator_instance_payload, decimator_payload, filter_unused_columns)
 
-    def apply_decimator(
+    def _apply_decimator(
         self,
         dataframe: pd.DataFrame,
         x_column_name: t.Optional[str],
         y_column_name: str,
         z_column_name: str,
         payload: t.Dict[str, t.Any],
         is_copied: bool,
-    ):
+    ) -> t.Tuple:
         """NOT DOCUMENTED
         This function is executed whenever a decimator is applied to the data.
 
@@ -263,7 +262,7 @@ def apply_decimator(
         return self._df_apply_decimator(dataframe, x_column_name, y_column_name, z_column_name, payload, is_copied)
 
     @abstractmethod
-    def decimate(self, data: np.ndarray, payload: t.Dict[str, t.Any]) -> np.ndarray:
+    def _decimate(self, data: np.ndarray, payload: t.Dict[str, t.Any]) -> np.ndarray:
         """NOT DOCUMENTED
         Decimate the dataset.
 

taipy/gui/data/decimator/lttb.py

@@ -58,7 +58,7 @@ def _areas_of_triangles(a, bs, c):
         a_minus_bs = a - bs
         return 0.5 * abs((a[0] - c[0]) * (bs_minus_a[:, 1]) - (a_minus_bs[:, 0]) * (c[1] - a[1]))
 
-    def decimate(self, data: np.ndarray, payload: t.Dict[str, t.Any]) -> np.ndarray:
+    def _decimate(self, data: np.ndarray, payload: t.Dict[str, t.Any]) -> np.ndarray:
         n_out = self._n_out
         if n_out >= data.shape[0]:
             return np.full(len(data), True)

taipy/gui/data/decimator/minmax.py

@@ -51,7 +51,7 @@ def __init__(
         super().__init__(threshold, zoom)
         self._n_out = n_out // 2
 
-    def decimate(self, data: np.ndarray, payload: t.Dict[str, t.Any]) -> np.ndarray:
+    def _decimate(self, data: np.ndarray, payload: t.Dict[str, t.Any]) -> np.ndarray:
         if self._n_out >= data.shape[0]:
             return np.full(len(data), False)
         # Create a boolean mask

taipy/gui/data/decimator/rdp.py

@@ -60,7 +60,7 @@ def __init__(
         self._n_out = n_out
 
     @staticmethod
-    def dsquared_line_points(P1, P2, points):
+    def __dsquared_line_points(P1, P2, points):
         """
         Calculate only squared distance, only needed for comparison
         """
@@ -93,7 +93,7 @@ def __rdp_epsilon(data, epsilon: int):
             P1 = data[start]
             P2 = data[end]
             points = data[start + 1 : end]
-            dsq = RDP.dsquared_line_points(P1, P2, points)
+            dsq = RDP.__dsquared_line_points(P1, P2, points)
 
             mask_eps = dsq > epsilon**2
 
@@ -129,7 +129,7 @@ def __rdp_points(M, n_out):
             (start, end) = stack.pop()
             if end - start <= 1:
                 continue
-            dsq = RDP.dsquared_line_points(M[start], M[end], M[start + 1 : end])
+            dsq = RDP.__dsquared_line_points(M[start], M[end], M[start + 1 : end])
             max_dist_index = np.argmax(dsq) + start + 1
             weights[max_dist_index] = np.amax(dsq)
             stack.append((start, max_dist_index))
@@ -138,9 +138,9 @@ def __rdp_points(M, n_out):
 
         return weights >= maxTolerance
 
-    def decimate(self, data: np.ndarray, payload: t.Dict[str, t.Any]) -> np.ndarray:
+    def _decimate(self, data: np.ndarray, payload: t.Dict[str, t.Any]) -> np.ndarray:
         if self._epsilon:
             return RDP.__rdp_epsilon(data, self._epsilon)
         elif self._n_out:
             return RDP.__rdp_points(data, self._n_out)
-        raise RuntimeError("RDP Decimator failed to run. Fill in either 'epsilon' or 'n_out' value")
+        raise RuntimeError("RDP Decimator failed to run. One of 'epsilon' or 'n_out' values must be specified")

taipy/gui/data/decimator/scatter_decimator.py

@@ -59,7 +59,7 @@ def __init__(
         self._binning_ratio = binning_ratio if binning_ratio > 0 else 1
         self._max_overlap_points = max_overlap_points if max_overlap_points is not None else 3
 
-    def decimate(self, data: np.ndarray, payload: t.Dict[str, t.Any]) -> np.ndarray:
+    def _decimate(self, data: np.ndarray, payload: t.Dict[str, t.Any]) -> np.ndarray:
         n_rows = data.shape[0]
         mask = np.empty(n_rows, dtype=bool)
         width = payload.get("width", None)

taipy/gui/data/pandas_data_accessor.py

@@ -411,7 +411,7 @@ def __get_data(  # noqa: C901
                 if isinstance(decimator_instance, PropertyType.decimator.value):
                     # Run the on_decimate method -> check if the decimator should be applied
                     # -> apply the decimator
-                    decimated_df, is_decimator_applied, is_copied = decimator_instance.on_decimate(
+                    decimated_df, is_decimator_applied, is_copied = decimator_instance._on_decimate(
                         df, decimator_pl, decimator_payload, is_copied
                     )
                     # add decimated dataframe to the list of decimated