stickler

pvlib · shirubana · Sep 22, 2022 · Sep 22, 2022 · Sep 22, 2022 · Sep 22, 2022
commit 2de5a3716272da5b4e8c7ae20dbb762767f2b093
diff --git a/pvlib/iotools/sam.py b/pvlib/iotools/sam.py
@@ -2,52 +2,61 @@
 
 import pandas as pd
 
-def saveSAM_WeatherFile(data, metadata, savefile='SAM_WeatherFile.csv', standardSAM = True, includeminute=False):
+def saveSAM_WeatherFile(data, metadata, savefile='SAM_WeatherFile.csv', 
+                        standardSAM=True, includeminute=False):
     """
-    Saves a dataframe with weather data from pvlib format on SAM-friendly format.
+    Saves dataframe with weather data from pvlib format on SAM-friendly format.
 
     Parameters
     -----------
     data : pandas.DataFrame
-        timeseries data in PVLib format. Should be TZ converted (not UTC). Ideally it is one sequential year data; if not suggested to use standardSAM = False. 
+        timeseries data in PVLib format. Should be TZ converted (not UTC).
+        Ideally it is one sequential year data; if not suggested to use
+        standardSAM = False.
     metdata : dictionary
-        Dictionary with at least 'latitude', 'longitude', 'elevation', 'source', and 'TZ' for timezone.
+        Dictionary with 'latitude', 'longitude', 'elevation', 'source',
+        and 'TZ' for timezone.
     savefile : str
-        Name of file to save output as. 
+        Name of file to save output as.
     standardSAM : boolean
-        This checks the dataframe to avoid having a leap day, then averages it to SAM style (closed to the right), 
-        and fills the years so it starst on YEAR/1/1 0:0 and ends on YEAR/12/31 23:00.
+        This checks the dataframe to avoid having a leap day, then averages it
+        to SAM style (closed to the right),
+        and fills the years so it starst on YEAR/1/1 0:0 and ends on
+        YEAR/12/31 23:00.
     includeminute ; Bool
-        For hourly data, if SAM input does not have Minutes, it calculates the sun position 30 minutes 
-        prior to the hour (i.e. 12 timestamp means sun position at 11:30)
-        If minutes are included, it will calculate the sun position at the time of the timestamp (12:00 at 12:00)
+        For hourly data, if SAM input does not have Minutes, it calculates the
+        sun position 30 minutes prior to the hour (i.e. 12 timestamp means sun
+         position at 11:30).
+        If minutes are included, it will calculate the sun position at the time
+        of the timestamp (12:00 at 12:00)
         Set to true if resolution of data is sub-hourly.
 
     Returns
     -------
     Nothing, it just writes the file.
-    
+
     """
 
     def _is_leap_and_29Feb(s):
-        ''' Creates a mask to help remove Leap Years. Obtained from: 
-            https://stackoverflow.com/questions/34966422/remove-leap-year-day-from-pandas-dataframe/34966636
+        ''' Creates a mask to help remove Leap Years. Obtained from:
+            https://stackoverflow.com/questions/34966422/remove-leap-year-day-
+            from-pandas-dataframe/34966636
         '''
         return (s.index.year % 4 == 0) & \
                ((s.index.year % 100 != 0) | (s.index.year % 400 == 0)) & \
                (s.index.month == 2) & (s.index.day == 29)
 
     def _averageSAMStyle(df, interval='60T', closed='right', label='right'):
-        ''' Averages subhourly data into hourly data in SAM's expected format. 
+        ''' Averages subhourly data into hourly data in SAM's expected format.
         '''
         try:
-            df = df.resample(interval, closed=closed, label=label).mean() #
+            df = df.resample(interval, closed=closed, label=label).mean()
         except:
             print('Warning - unable to average')
         return df
 
     def _fillYearSAMStyle(df, freq='60T'):
-        ''' Fills year 
+        ''' Fills year
         '''
         # add zeros for the rest of the year
         if freq is None:
@@ -59,38 +68,41 @@ def _fillYearSAMStyle(df, freq='60T'):
         # idx = df.index
         # apply correct TZ info (if applicable)
         tzinfo = df.index.tzinfo
-        starttime = pd.to_datetime('%s-%s-%s %s:%s' % (df.index.year[0],1,1,0,0 ) ).tz_localize(tzinfo)
-        endtime = pd.to_datetime('%s-%s-%s %s:%s' % (df.index.year[-1],12,31,23,60-int(freq[:-1])) ).tz_localize(tzinfo)
+        starttime = pd.to_datetime('%s-%s-%s %s:%s' % (df.index.year[0], 1, 1,
+                                                      0, 0)).tz_localize(tzinfo)
+        endtime = pd.to_datetime('%s-%s-%s %s:%s' % (df.index.year[-1], 12, 31,
+                                                     23, 60-int(freq[:-1]))
+                                                     ).tz_localize(tzinfo)
 
         df2 = _averageSAMStyle(df, freq)
         df2.iloc[0] = 0  # set first datapt to zero to forward fill w zeros
         df2.iloc[-1] = 0  # set last datapt to zero to forward fill w zeros
         df2.loc[starttime] = 0
         df2.loc[endtime] = 0
-        df2 = df2.resample(freq).ffill() 
+        df2 = df2.resample(freq).ffill()
         return df2
 
-
-    # Modify this to cut into different years. Right now handles partial year and sub-hourly interval.
+    # Modify this to cut into different years. Right now handles partial year
+    # and sub-hourly interval.
     if standardSAM:
-        filterdatesLeapYear = ~(_is_leap_and_29Feb(data)) 
+        filterdatesLeapYear = ~(_is_leap_and_29Feb(data))
         data = data[filterdatesLeapYear]
         data = _fillYearSAMStyle(data)
 
-
     # metadata
-    latitude=metadata['latitude']
-    longitude=metadata['longitude']
-    elevation=metadata['elevation']
+    latitude = metadata['latitude']
+    longitude = metadata['longitude']
+    elevation = metadata['elevation']
     timezone_offset = metadata['TZ']
     source = metadata['source']
-
-    # make a header        
-    header = '\n'.join(
-            [ 'Source,Latitude,Longitude,Time Zone,Elevation',
-            source + ',' + str(latitude) + ',' + str(longitude) + ',' + str(timezone_offset) + ',' + str(elevation)]) + '\n'
 
-    savedata = pd.DataFrame({'Year':data.index.year, 'Month':data.index.month, 'Day':data.index.day,
+    # make a header
+    header = '\n'.join([ 'Source,Latitude,Longitude,Time Zone,Elevation',
+                        source + ',' + str(latitude) + ',' + str(longitude) 
+                        + ',' + str(timezone_offset) + ',' + str(elevation)])+'\n'
+
+    savedata = pd.DataFrame({'Year':data.index.year, 'Month':data.index.month,
+                             'Day':data.index.day,
                              'Hour':data.index.hour})
 
     if includeminute:
@@ -100,24 +112,24 @@ def _fillYearSAMStyle(df, freq='60T'):
     temp_amb = list(data.temp_air)
     savedata['Wspd'] = windspeed
     savedata['Tdry'] = temp_amb
-    
+
     if 'dni' in data:
         dni = list(data.dni)
         savedata['DHI'] = dni
-    
+
     if 'dhi' in data:
         dhi = list(data.dhi)
         savedata['DNI'] = dhi
-                            
+
     if 'ghi' in data:
         ghi = list(data.ghi)
         savedata['GHI'] = ghi
-    
-    if 'poa' in data: # This is a nifty function of SAM for using field measured POA irradiance!
+
+    if 'poa' in data:
         poa = list(data.poa)
         savedata['POA'] = poa
-        
-    if 'albedo' in data: 
+
+    if 'albedo' in data:
         albedo = list(data.albedo)
         savedata['Albedo'] = albedo
 
@@ -133,18 +145,18 @@ def _fillYearSAMStyle(df, freq='60T'):
 
 def tz_convert(df, tz_convert_val, metadata=None):
     """
-    Support function to convert metdata
E78F
 to a different local timezone. Particularly for 
-    GIS weather files which are returned in UTC by default.
-    
+    Support function to convert metdata to a different local timezone. 
+    Particularly for GIS weather files which are returned in UTC by default.
+
     Parameters
     ----------
     df : DataFrame
         A dataframe in UTC timezone
     tz_convert_val : int
-        Convert timezone to this fixed value, following ISO standard 
+        Convert timezone to this fixed value, following ISO standard
         (negative values indicating West of UTC.)
-    Returns: metdata, metadata  
-    
+    Returns: metdata, metadata
+
 
     Returns
     -------

diff --git a/pvlib/tests/iotools/test_sam.py b/pvlib/tests/iotools/test_sam.py
@@ -17,12 +17,16 @@
 testfile_radiation_csv = r'C:\Users\sayala\Documents\GitHub\pvlib-python\pvlib\data\pvgis_hourly_Timeseries_45.000_8.000_SA_30deg_0deg_2016_2016.csv'
 
 def test_saveSAM_WeatherFile():
-    data, inputs, metadata = read_pvgis_hourly(testfile_radiation_csv, map_variables=True)#, pvgis_format=pvgis_format)
-    metadata = {'latitude': inputs['latitude'],
-           'longitude': inputs['longitude'],
-           'elevation': inputs['elevation'],
+    data, months_selected, inputs, metadata = get_pvgis_tmy(latitude=33, longitude=-110, map_variables=True)
+    #  Reading local file doesn't work read_pvgis_hourly returns different 
+    # keys than get_pvgis_tmy when map=variables = True.  Opened issue for that
+#    data, inputs, metadata = read_pvgis_hourly(testfile_radiation_csv, 
+#                                               map_variables=True)
+    metadata = {'latitude': inputs['location']['latitude'],
+           'longitude': inputs['location']['longitude'],
+           'elevation': inputs['location']['elevation'],
            'source': 'User-generated'}
-    metadata['TZ'] = -7        
+    metadata['TZ'] = -7
     data = tz_convert(data, tz_convert_val=metadata['TZ'])
     coerce_year=2021  # read_pvgis_hourly does not coerce_year, so doing it here.
     data.index = data.index.map(lambda dt: dt.replace(year=coerce_year))