Merge pull request #44 from pnnl/fix_preprocess_api_init

Fix preprocessing API initialization

constrain/api/data_processing.py

@@ -30,7 +30,6 @@ def __init__(
             data_source (str): Data source name. Use `EnergyPlus` or `Other`.
             timestamp_column_name (str): Name of the column header that contains the time series timestamps.
         """
-
         self.data = None
 
         if data_path is None:
@@ -44,13 +43,13 @@ def __init__(
         # check if data file exists
         if os.path.isfile(data_path):
             try:
-                if data_source == "EnergyPlus":
+                if data_source.lower() == "energyplus":
                     # Use CSVReader to parse EnergyPlus timestamps
                     data = CSVReader(csv_file=data_path).getseries()
                     data = DateTimeEP(data, 2000).transform()
                     data.drop("Date/Time", inplace=True, axis=1)
 
-                elif data_source == "Other":
+                elif data_source.lower() == "bms":
                     if timestamp_column_name is None:
                         logging.error(
                             "timestamp_column_name is required when data_source = 'Other'"
@@ -65,7 +64,7 @@ def __init__(
                         return None
                     data.set_index(timestamp_column_name, inplace=True)
                     try:
-                        data = pd.to_datetime(data.index)
+                        data.index = pd.to_datetime(data.index)
                     except:
                         logging.error(
                             f"The data in {timestamp_column_name} could not be converted to Python datetime object. Make sure that the data is consistent defined as a set of date strings."

docs/source/conf.py

@@ -33,6 +33,8 @@
 # The full version, including alpha/beta/rc tags
 release = "0.4.0"
 
+# Document class docstrings
+autoclass_content = "both"
 
 # -- General configuration ---------------------------------------------------
 

tests/api/test_data_processing.py

@@ -1,4 +1,4 @@
-import unittest, sys, datetime, copy
+import unittest, sys, datetime, copy, pandas
 
 import matplotlib
 
@@ -76,18 +76,20 @@ def test_ep_data_file(self):
         t = DataProcessing(data_path=filep, data_source="EnergyPlus")
         assert len(t.data) == 2
 
-    def test_other_data_file(self):
+    def test_BMS_data_file(self):
         filep = "./tests/api/data/data_non_ep_head.csv"
         dp = DataProcessing(
-            data_path=filep, data_source="Other", timestamp_column_name="Date Time"
+            data_path=filep, data_source="BMS", timestamp_column_name="Date Time"
         )
         assert len(dp.data) == 2
+        assert len(dp.data.columns) == 106
+        assert isinstance(dp.data.index, pandas.DatetimeIndex)
 
     def test_datafile_error_parsing(self):
         with self.assertLogs() as logobs:
             filep = "./tests/api/data/data_err_parse.csv"
             dp = DataProcessing(
-                data_path=filep, data_source="Other", timestamp_column_name="Date/Time"
+                data_path=filep, data_source="BMS", timestamp_column_name="Date/Time"
             )
             self.assertEqual(
                 f"ERROR:root:The data in Date/Time could not be converted to Python datetime object. Make sure that the data is consistent defined as a set of date strings.",
@@ -97,9 +99,9 @@ def test_datafile_error_parsing(self):
     def test_datafile_missing_datetimecol(self):
         with self.assertLogs() as logobs:
             filep = "./tests/api/data/data_non_ep_head.csv"
-            dp = DataProcessing(data_path=filep, data_source="Other")
+            dp = DataProcessing(data_path=filep, data_source="BMS")
             self.assertEqual(
-                "ERROR:root:timestamp_column_name is required when data_source = 'Other'",
+                "ERROR:root:timestamp_column_name is required when data_source = 'BMS'",
                 logobs.output[0],
             )