working on weather data

energy_demand/read_write/data_loader.py

@@ -892,6 +892,17 @@ def load_temp_data(local_paths, weather_yrs_scenario, save_fig=False):
     temp_data : dict
         Temperatures
     """
+
+    '''
+    # Get path to weather_yr
+
+    # Get all station_ids with coordinates
+
+
+    # Get all min, max weather data
+    '''
+
+
     temp_data_short = defaultdict(dict)
     weather_stations_with_data = defaultdict(dict)
 

energy_demand/scripts/weather_scenario.py

@@ -9,16 +9,146 @@
 http://joehamman.com/2013/10/12/plotting-netCDF-data-with-Python/
 
 """
+import csv
 import pandas as pd
 import pytemperature
 import xarray as xr
 
-tasmax = "C:/_WEATHERDATA/nf-2020/2020/m00m/daily/WAH_m00m_tasmax_daily_g2_2020.nc"
-tasmin = "C:/_WEATHERDATA/nf-2020/2020/m00m/daily/WAH_m00m_tasmin_daily_g2_2020.nc"
+def get_temp_data_from_nc(path_nc_file, attribute_to_keep):
+    """Open c file, convert it into dataframe,
+    clean it and extract attribute data
+    """
+    # Open as xarray
+    x_array_data = xr.open_dataset(path_nc_file)
+
+    # Convert to dataframe
+    df = x_array_data.to_dataframe()
+
+    # Add index as columns
+    df = df.reset_index()
+
+    # Drop unnecessary columns
+    df = df.drop(['time_bnds', 'rotated_latitude_longitude', 'rlat', 'rlon', 'bnds', 'height'], axis=1)
+
+    # Drop all missing value
+    df = df.dropna(subset=[attribute_to_keep])
+
+    return df
+
+def convert_to_celcius(df, attribute_to_convert):
+    """# Convert Kelvin to Celsius (# Kelvin to Celsius)
+    """
+    df[attribute_to_convert] = df[attribute_to_convert].apply(pytemperature.k2c)
+    return df
+
+def extend_360_day_to_365(df, attribute):
+    """Add evenly distributed days across the 360 days
+
+    Return
+    ------
+    out_list : list
+        Returns list with days with the following attributes
+        e.g. [[lat, long, yearday365, value]]
+    """
+    # Copy the following position of day in the 360 year
+    days_to_duplicate = [60, 120, 180, 240, 300]
+
+    #columns = df.columns
+    #df_365 = pd.DataFrame(columns=columns)
+
+    out_list = []
+    # Copy every sithieth day over the year and duplicate it. Do this five times --> get from 360 to 365 days
+    vals = []
+    day_cnt, day_cnt_365 = 0, 0
+
+    for index, row in df.iterrows():
+        day_cnt += 1
+        # Create row into dataframe
+        #df_row = pd.DataFrame([list(row)], columns=columns)
+
+        # Copy extra day
+        if day_cnt in days_to_duplicate:
+            #df_row['time'] = day_cnt_365 # Add yearday
+            #df_365 = df_365.append(df_row)
+            out_list.append([row['lat'], row['lon'], day_cnt_365, row[attribute]])
+            day_cnt_365 += 1
+
+        # Copy day
+        day_cnt_365 += 1
+        #df_row['time'] = day_cnt_365 # Add yearday
+        #df_365 = df_365.append(df_row)
+        out_list.append([row['lat'], row['lon'], day_cnt_365, row[attribute]])
+
+        # Reset counter if next weather station
+        if day_cnt == 360:
+            day_cnt = 0
+            day_cnt_365 = 0
+
+    return out_list
+
+def write_wether_data():
+    """Write weather data to array
+    """
+    #station_coordinates = 
+    #station_dat = 
+
+    nr_of_entries = len(min_vals)
+    cnt = 0
+    for i in range(nr_of_entries):
+        print("o " + str(i))
+        min_val = max_vals[cnt] 
+        max_val = min_vals[cnt] 
+        lon = longs[cnt]
+        lat = lats[cnt]
+
+        if cnt == 0:
+            station_name = "station_{}_{}.csv".format(lon, lat)
+            station = pd.DataFrame(columns=list('tasmax', 'tasmin'))
+            path_station = "C:/_scrap/{}".format(station_name)
+
+        station_row = pd.DataFrame([[lon, lat, min_val, max_val]], columns=list( 'tasmax', 'tasmin'))
+
+        station.append(station_row)
+
+        if cnt == 364:
+            # Write station
+            station.to_csv(station, path_station)
+            station_id_positions.append([station_name, lon, lat])
+            cnt = 0
+
+        cnt +=1
+
+    return station_coordinates, station_data
+
+
+path_tasmax = "C:/_WEATHERDATA/nf-2020/2020/m00m/daily/WAH_m00m_tasmax_daily_g2_2020.nc"
+path_tasmin = "C:/_WEATHERDATA/nf-2020/2020/m00m/daily/WAH_m00m_tasmin_daily_g2_2020.nc"
 
 out_path = "C:/_scrap/test.csv"
 out_path_station_names = "C:/_scrap/station_position.csv"
-ds_max = xr.open_dataset(tasmax)
+
+# Load data
+df_max = get_temp_data_from_nc(path_tasmax, 'tasmax')
+df_min = get_temp_data_from_nc(path_tasmin, 'tasmin')
+
+# Convert Kelvin to Celsius (# Kelvin to Celsius)
+df_max['tasmax'] = convert_to_celcius(df_max, 'tasmax')
+df_min['tasmin'] = convert_to_celcius(df_min, 'tasmin')
+
+# Convert 360 day to 365 days
+list_max = extend_360_day_to_365(df_max, 'tasmax')
+list_min = extend_360_day_to_365(df_min, 'tasmin')
+
+# Write out single weather stations as numpy array
+
+# Write out coordinates of weather data to csv
+
+
+
+
+
+raise Exception("ff")
+'''ds_max = xr.open_dataset(tasmax)
 ds_min = xr.open_dataset(tasmin)
 
 df_max = ds_max.to_dataframe()
@@ -39,8 +169,8 @@
 
 # Convert Kelvin to Celsius (# Kelvin to Celsius)
 df_max['tasmax'] = df_max['tasmax'].apply(pytemperature.k2c)
-df_min['tasmin'] = df_min['tasmin'].apply(pytemperature.k2c)
-
+df_min['tasmin'] = df_min['tasmin'].apply(pytemperature.k2c)'''
+'''
 # Add evenly distributed days
 # Copy every sithieth day over the year and duplicate it. Do this five times --> get from 360 to 365 days
 min_vals = []
@@ -86,7 +216,7 @@
 #print(max_vals)
 print(len(min_vals))
 print(len(max_vals))
-
+'''
 nr_of_entries = len(min_vals)
 cnt = 0
 for i in range(nr_of_entries):