added create_dataset.py

src/methane_super_emitters/convert.py

@@ -10,6 +10,7 @@
 import matplotlib.pyplot as plt
 import click
 import numpy.ma as ma
+import datetime
 
 @click.command()
 @click.option('-i', '--input-file', help='Input netCDF file')
@@ -36,7 +37,9 @@ def main(input_file, output_file):
             methane_matrix[sl][gp] = m
             lat_matrix[sl][gp] = lt
             lon_matrix[sl][gp] = ln
-            time_matrix[sl][gp] = time
+            if time is not None:
+                time_matrix[sl][gp] = datetime.datetime(time.data[0], time.data[1], time.data[2],
+                                                        time.data[3], time.data[4], time.data[5])
 
     np.savez(output_file, methane=methane_matrix, lat=lat_matrix, lon=lon_matrix, time=time_matrix)
 

src/methane_super_emitters/create_dataset.py

@@ -0,0 +1,54 @@
+import click
+import numpy as np
+import datetime
+
+def check_if_inside(csv_line, lat_window, lon_window, time_window):
+    if np.count_nonzero(lat_window == -1000) == (lat_window.shape[0] * lat_window.shape[1]):
+        return False
+    csv_line = csv_line.strip().split(',')
+    csv_year = int(csv_line[0][0:4])
+    csv_month = int(csv_line[0][4:6])
+    csv_day = int(csv_line[0][6:8])
+    csv_time = csv_line[1].split(':')
+    csv_hour = int(csv_time[0])
+    csv_minute = int(csv_time[1])
+    csv_second = int(csv_time[2])
+    csv_datetime = datetime.datetime(csv_year, csv_month, csv_day, csv_hour, csv_minute, csv_second)
+    csv_lat = float(csv_line[2])
+    csv_lon = float(csv_line[3])
+    mask = lat_window != -1000
+    min_datetime = time_window[mask].min()
+    max_datetime = time_window[mask].max()
+    min_lat = lat_window[mask].min()
+    max_lat = lat_window[mask].max()
+    min_lon = lon_window[mask].min()
+    max_lon = lon_window[mask].max()
+    return ((min_datetime <= csv_datetime <= max_datetime) and
+            (min_lat <= csv_lat <= max_lat) and
+            (min_lon <= csv_lon <= max_lon))
+
+@click.command()
+@click.option('-i', '--input-file', help='Input CSV with super-emitter locations')
+@click.option('-m', '--matrix-file', help='Input NPZ file with methane data from TROPOMI')
+def main(input_file, matrix_file):
+    methane_data = np.load(matrix_file, allow_pickle=True)
+    with open(input_file, 'r') as fd:
+        data = fd.readlines()[1:]
+    methane_matrix = methane_data['methane']
+    lat_matrix = methane_data['lat']
+    lon_matrix = methane_data['lon']
+    time_matrix = methane_data['time']
+    rows, cols = methane_matrix.shape
+    for row in range(0, rows, 16):
+        for col in range(0, cols, 16):
+            if row + 32 < rows and col + 32 < cols:
+                methane_window = methane_matrix[row:row + 32][:, col:col + 32]
+                lat_window = lat_matrix[row:row + 32][:, col:col + 32]
+                lon_window = lon_matrix[row:row + 32][:, col:col + 32]
+                time_window = time_matrix[row:row + 32][:, col:col + 32]
+                for csv_line in data:
+                    if check_if_inside(csv_line, lat_window, lon_window, time_window):
+                        print("FOUND!")
+
+if __name__ == '__main__':
+    main()