initial io module prototype

pvlib/iotools/api.py

@@ -0,0 +1,2 @@
+from pvlib.iotools.tmy import readtmy2, readtmy3
+from pvlib.ioiotools.maccrad import read_maccrad

pvlib/iotools/iotools.py

@@ -0,0 +1,33 @@
+import pytz
+# How to get a time zone from a location using 
+# latitude and longitude coordinates?
+# http://stackoverflow.com/a/16086964
+# upstream: https://github.com/MrMinimal64/timezonefinder
+from timezonefinder import TimezoneFinder
+
+def get_loc_latlon(lat, lon):
+    """Returns the timezone for a coordinate pair.
+    """
+
+
+    tf = TimezoneFinder()
+    tz = tf.timezone_at(lng=lon, lat=lat)
+
+    return tz
+
+def localise_df(df_notz, tz_source_str='UTC', tz_target_str=None):
+    """
+    localises a pandas.DataFrame (df) to the target time zone of the pvlib-Location    
+
+    Assumes that the input df does not have a timezone
+
+    """
+
+    tz_source = pytz.timezone(tz_source_str)
+    tz_target = pytz.timezone(tz_target_str)
+
+    df_tz_source = df_notz.tz_localize(tz_source)
+    df_tz_target = df_tz_source.tz_convert(tz_target)
+
+
+    return df_tz_target

pvlib/iotools/maccrad.py

@@ -0,0 +1,152 @@
+# standard library imports 
+import logging
+
+# related third party imports
+import pandas as pd
+import pytz
+
+# local application/library specific imports 
+from pvlib.location import Location
+from pvlib.io.iotools import get_loc_latlon, localise_df
+
+# required dateconverters
+def dtp_soda_pro_macc_rad(date):
+    """
+    datetime converter for MACC-RAD data and others
+    """
+    datetime_stamp = date.split('/')[0]
+
+    return datetime_stamp
+
+#XXX read metadata / header
+
+def read_maccrad_metadata(file_csv, name='maccrad'):
+    """
+    Read metadata from commented lines of the file
+    * coordinates
+    * altitude
+
+    Retrieve timezone
+    """
+    if not name:
+        name = file_csv.split('.')[0]
+
+    ## if file is on local drive
+    f = open(file_csv)
+    for line in f:
+        if "Title" in line:
+            title = line.split(':')[1].split('(')[0].strip()
+            name = title
+        if "Latitude" in line:
+    #    print (line)
+    #    if line.startswith( "# Latitude"):
+            lat_line = line
+            lat = float(lat_line.split(':')[1])
+    #        lat = float(line.split(':')[1])
+        if "Longitude" in line:
+    #    if line.startswith( "# Longitude"):
+            lon_line = line
+            lon = float(lon_line.split(':')[1])
+            lon = float(line.split(':')[1])
+    #    if line.startswith( "# Altitude"):
+        if "Altitude" in line:
+            alt_line = line
+            alt = float(alt_line.split(':')[1])
+    #        alt = float(line.split(':')[1])
+        if "Time reference" in line:
+            if "Universal time (UT)" in line:
+                tz_raw = 'UTC'
+            else:
+                logging.debug('No metadata on timezone found in input file')
+                logging.debug('Assuming UTC as default timezone')
+                tz_raw = 'UTC'
+
+    tz_loc = get_loc_latlon(lat, lon)
+
+
+
+    location = Location(lat, lon, name=name, altitude=alt,
+                           tz=tz_loc)
+
+    return tz_raw, location
+
+
+def maccrad_df_to_pvlib(df_raw, tz_raw, loc, localise=True):
+    """Change some properties of the dataframe to be more compliant with pvlib
+
+    * localisation
+    * column renaming
+    * setting dataframe name description according to datasource
+
+    """
+
+    if localise:
+        # timezone localisations
+        df_pvlib = localise_df(df_raw, tz_source_str=tz_raw, 
+                               tz_target_str=loc.tz)
+    # column renaming
+    df_pvlib.index.name = 'datetime'
+
+    # name the dataframe according to data source 
+    df_pvlib.df_name = loc.name
+
+    return df_pvlib
+
+#def maccrad_df_to_pvlib(df):
+#    
+#    
+#    pass
+
+
+#XXX read data
+def read_maccrad(file_csv, loc_name=None, skiprows=40, output='all'):
+    """
+    Read MACC-RAD current format for files into a pvlib-ready dataframe
+
+    Parameters
+    ----------
+    file_csv : a csv file corresponding to the reader format   
+    skiprows : skiprows as in pandas.io.read_csv
+        The example files require skipping 40 rows.
+    output : all / loc / df_pvlib        
+        df_raw returns only the a pandas.DataFrame using the raw data from the 
+                        file (this can be helpful for debugging and comparison 
+                        with restuls obtained with other programs, e.g. 
+                        spreadsheet)
+        all returns df_raw, returns a pandas.DataFrame reformatted to match the 
+                         `variable naming convention <variables_style_rules>`
+                         for pvliball outputs, and a location created from 
+                         the metadata in raw input file header
+                         as tuple
+
+
+    """
+    df_raw = pd.read_csv(file_csv, sep=';', skiprows=skiprows, header=0,
+                      index_col=0, parse_dates=True,
+                      date_parser=dtp_soda_pro_macc_rad)
+#TODO: add loc_name
+#TODO: add reformat needs loc!
+#TODO: add simplify output options raw or all
+    print (output)
+    if output == 'df_raw':
+        res = df_raw
+    if output == 'all':
+        tz_raw, loc = read_maccrad_metadata(file_csv)
+        loc.name = (loc.name + ' @ ' + 'lat (deg. N), lon (deg. E): ' + 
+                    str(loc.latitude) + ', ' + str(loc.longitude))
+        df_pvlib = maccrad_df_to_pvlib(df_raw, tz_raw, loc, localise=True)
+#        res = df_pvlib
+        res = (df_raw, df_pvlib, loc)
+#    if output == 'loc':
+#        
+#        res = loc, df
+#    if output == 'all':
+#        # not calculated outside conditional to reduce overhead of metadata 
+#        # reading if not desired
+#        loc = read_maccrad_metadata(file_csv)
+#        res = (df_raw, df_pvlib, loc)
+
+
+    return res
+
+