Chapter 2: temp forecast using pyowm

chapter2/solutions/nyc_forecast_owm.py

@@ -1,24 +1,39 @@
 '''
-Retrieve a day's forecast of New York City and create a graph
+Retrieve three days' forecast of New York City and create a graph
 Using https://github.com/csparpa/pyowm
 '''
 import matplotlib.pyplot as plt
 from pyowm import OWM
+import pytz
 
-owm = OWM()
+owm = OWM('<token>')
 
 def get_forecast(city):
-    fc = owm.daily_forecast(city, limit=1)
+    # https://github.com/csparpa/pyowm/issues/266
+    fc = owm.three_hours_forecast(city)
     f = fc.get_forecast()
-    w = f.get_weathers()[0]
-    forecast_temp = w.get_temperature('celsius')
-    day_intervals = ['morn', 'day', 'eve', 'night']
+
+    # three_hours_forecast() returns 5 day forecast at a granularity
+    # of three hours
+    # we plot them all
+    weathers = f.get_weathers()
+
+
+    data_points = ['temp', 'temp_max', 'temp_min', 'temp_kf']
     temp = []
-    for timeofday in day_intervals:
-        temp.append(forecast_temp[point])
-    x = range(1, len(day_intervals)+1)
+    date_time = []
+    for w in weathers:
+        forecast_temp = w.get_temperature('celsius')
+        utc_dt = datetime.utcfromtimestamp(w.get_reference_time()).replace(tzinfo=pytz.utc)
+        tz = pytz.timezone('America/New_York')
+        dt = utc_dt.astimezone(tz)
+
+        # print it
+        date_time.append(dt.strftime('%Y-%m-%d %H:%M:%S %Z%z'))
+        temp.append(forecast_temp['temp'])
+    x = range(1, len(temp)+1)
     plt.plot(x, temp, 'o-')
-    plt.xticks(x, day_intervals)
+    plt.xticks(x, date_time)
     plt.show()
 
 if __name__ == '__main__':