Merge pull request #375 from oist/feature/datamodules

Add: separate datamodules

optinist/api/dataclass/bar.py

@@ -0,0 +1,35 @@
+import numpy as np
+import pandas as pd
+
+from optinist.api.dataclass.base import BaseData
+from optinist.api.utils.filepath_creater import join_filepath
+from optinist.api.utils.json_writer import JsonWriter
+
+
+class BarData(BaseData):
+    def __init__(self, data, index=None, file_name='bar'):
+        super().__init__(file_name)
+        data = np.array(data)
+
+        assert data.ndim <= 2, 'Bar Dimension Error'
+
+        if data.ndim == 1:
+            data = data[np.newaxis]
+
+        assert data.ndim == 2, 'Bar Dimesion is not 2'
+
+        self.data = data
+
+        # indexを指定
+        if index is not None:
+            self.index = index
+        else:
+            self.index = np.arange(len(self.data))
+
+    def save_json(self, json_dir):
+        self.json_path = join_filepath([json_dir, f"{self.file_name}.json"])
+        df = pd.DataFrame(
+            self.data,
+            index=self.index,
+        )
+        JsonWriter.write_as_split(self.json_path, df)

optinist/api/dataclass/base.py

@@ -0,0 +1,13 @@
+import gc
+
+
+class BaseData:
+    def __init__(self, file_name):
+        self.file_name = file_name
+
+    def save_json(self, json_dir):
+        pass
+
+    def __del__(self):
+        del self
+        gc.collect()

optinist/api/dataclass/csv.py

@@ -0,0 +1,38 @@
+import numpy as np
+import pandas as pd
+
+from optinist.api.dataclass.base import BaseData
+
+from optinist.api.utils.filepath_creater import create_directory, join_filepath
+from optinist.api.utils.json_writer import JsonWriter
+
+
+class CsvData(BaseData):
+    def __init__(self, data, params, file_name='csv'):
+        super().__init__(file_name)
+
+        if isinstance(data, str):
+            self.data = pd.read_csv(data, header=None).values
+
+            if params["transpose"]:
+                self.data = self.data.T
+
+            if params["setHeader"] is not None:
+                header = params["setHeader"]
+                self.data = self.data[header:]
+        else:
+            self.data = np.array(data)
+
+        if self.data.ndim == 1:
+            self.data = self.data[np.newaxis, :]
+
+    def save_json(self, json_dir):
+        # timeseriesだけはdirを返す
+        self.json_path = join_filepath([json_dir, self.file_name])
+        create_directory(self.json_path)
+
+        for i, data in enumerate(self.data):
+            JsonWriter.write_as_split(
+                join_filepath([self.json_path, f'{str(i)}.json']),
+                data
+            )