[REF-6255] generic: adding a Generic engine section with some misc components

o Adding file_to_dataframe
o Adding dataframe_stats

Author: lior-amar

components/Python/Generic/dataframe_stats/__init__.py

@@ -0,0 +1 @@
+

components/Python/Generic/dataframe_stats/component.json

@@ -0,0 +1,57 @@
+{
+  "version": 1,
+  "engineType": "Generic",
+  "language": "Python",
+  "userStandalone": false,
+  "name": "dataframe_stats",
+  "label": "DataFrame Statistics",
+  "description": "Generate statistics on the input dataframe.",
+  "program": "dataframe_stats.py",
+  "componentClass": "MCenterStatsComponentAdapter",
+  "modelBehavior": "Auxiliary",
+  "useMLOps": true,
+  "inputInfo": [
+    {
+      "description": "In Pandas Dataframe",
+      "label": "dataframe",
+      "defaultComponent": "",
+      "type": "dataframe",
+      "group": "data"
+    }
+  ],
+  "outputInfo": [
+    {
+      "description": "Out Pandas Dataframe",
+      "label": "dataframe",
+      "defaultComponent": "",
+      "type": "dataframe",
+      "group": "data"
+    }
+  ],
+  "group": "FeatureEng",
+  "arguments": [
+    {
+      "key": "dataframe_is",
+      "label": "Dataframe Is",
+      "description": "What is this dataframe represents",
+      "type": "string",
+      "uiType": "select",
+      "options": [
+        {
+          "label": "Input Data",
+          "value": "input_data"
+        },
+        {
+          "label": "Categorical Predictions Probabilities",
+          "value": "categorical_predictions_probabilities"
+        },
+        {
+          "label": "Other",
+          "value": "other"
+        }
+      ],
+      "defaultValue": "input_data",
+      "optional": false
+    }
+  ]
+}

components/Python/Generic/dataframe_stats/dataframe_stats.py

@@ -0,0 +1,44 @@
+from __future__ import print_function
+
+from parallelm.components import ConnectableComponent
+from parallelm.mlops import mlops as mlops
+from parallelm.mlops.stats.bar_graph import BarGraph
+
+
+class MCenterStatsComponentAdapter(ConnectableComponent):
+    def __init__(self, engine):
+        super(self.__class__, self).__init__(engine)
+
+    def _materialize(self, parent_data_objs, user_data):
+        mlops.init()
+
+        df = parent_data_objs[0]
+        dataframe_is = self._params.get("dataframe_is", "input_data")
+
+        if dataframe_is == "input_data":
+            self._handle_input_data(df)
+        elif dataframe_is == "categorical_predictions_probabilities":
+            self._handle_categorical_predictions(df)
+        elif dataframe_is == "other":
+            pass
+        else:
+            self._logger("Error: argument value is not supported: {}".format(dataframe_is))
+
+        mlops.done()
+
+        return[df]
+
+    def _handle_input_data(self, df):
+        mlops.set_data_distribution_stat(df)
+
+    def _handle_categorical_predictions(self, df):
+
+        df_max_col = df.idxmax(axis=1)
+        series_value_count = df_max_col.value_counts(normalize=True)
+
+        col_values = []
+        for col in df.columns:
+            col_values.append(series_value_count.at[col])
+
+        bg = BarGraph().name("Categorical Prediction Distribution").cols(list(df.columns)).data(col_values)
+        mlops.set_stat(bg)

components/Python/Generic/file_to_dataframe/__init__.py

@@ -0,0 +1 @@
+

components/Python/Generic/file_to_dataframe/component.json

@@ -0,0 +1,38 @@
+{
+  "version": 1,
+  "engineType": "Generic",
+  "language": "Python",
+  "userStandalone": false,
+  "name": "file_to_dataframe",
+  "label": "Source File to DataFrame",
+  "program": "file_to_dataframe.py",
+  "componentClass": "MCenterComponentAdapter",
+  "modelBehavior": "Auxiliary",
+  "useMLOps": true,
+  "inputInfo": [{
+    "description": "File to read contents",
+    "label": "filename",
+    "defaultComponent": "",
+    "type": "str",
+    "group": "data"
+  }],
+  "outputInfo": [
+    {
+      "description": "Pandas Dataframe",
+      "label": "dataframe",
+      "defaultComponent": "",
+      "type": "dataframe",
+      "group": "data"
+    }
+  ],
+  "group": "Connectors",
+  "arguments": [
+    {
+      "key": "filename",
+      "label": "Dataset file to read",
+      "type": "str",
+      "description": "File to use for loading DataSet into DataFrame",
+      "optional": true
+    }
+  ]
+}

components/Python/Generic/file_to_dataframe/file_to_dataframe.py

@@ -0,0 +1,38 @@
+from __future__ import print_function
+
+import sys
+import os
+import pandas
+
+from parallelm.components import ConnectableComponent
+
+
+class MCenterComponentAdapter(ConnectableComponent):
+    """
+    Adapter for read_file_to_df
+    """
+
+    def __init__(self, engine):
+        super(self.__class__, self).__init__(engine)
+
+    def _materialize(self, parent_data_objs, user_data):
+        if len(parent_data_objs) is not 0:
+            file_path = str(parent_data_objs[0])
+        else:
+            file_path = self._params.get('filename')
+
+        self._logger.info("file: {}".format(file_path))
+        df = self.read_file_to_df(file_path)
+        return [df]
+
+    def read_file_to_df(self, filepath):
+        """
+        Read file and return DataFrame
+        """
+
+        if not os.path.exists(filepath):
+            self._logger.info("stderr- failed to find {}".format(filepath), file=sys.stderr)
+            raise Exception("file path does not exist: {}".format(filepath))
+
+        df = pandas.read_csv(filepath)
+        return df