def _weather_data_extraction(lat, lon):
start = datetime(2015, 1, 1)
end = datetime(2021, 1, 16)
stations = Stations()
stations = stations.nearby(lat, lon)
stations = stations.inventory('daily', (start, end))
station = stations.fetch(1)
data = Daily(station, start, end)
data = data.fetch()
return data
def get_historical_data(self):
"""
returns: Dataframe of historical average temperature for a given city
within the specified time period
"""
lat, lon = self.get_lat_and_lon()
city = Point(lat, lon)
data = Daily(city, self.start_date, self.end_date)
data = data.fetch()
data = data.filter(['tavg']) # select average temperature column
data = data.fillna(method='ffill') # fill missing values
# get a specific day
#day = data.iloc[lambda x: x.index == '2014-06-23']
return data
from datetime import datetime
from meteostat import Point, Daily
import seaborn as sns
import matplotlib.pyplot as plt
import matplotlib as mpl
import pandas as pd
start = datetime(2021, 1, 1)
end = datetime(2021, 12, 31)
oslo = Point(59.9139, 10.7522)
od = Daily(oslo, start, end)
od = od.fetch()
print(od.isnull().sum())
df = od.drop(['snow', 'wdir', 'wspd', 'wpgt', 'pres', 'tsun'], axis=1)
df = df.dropna()
print(df)
print(df.isna().sum())
corr = df.corr()
print(corr)
sns.heatmap(corr, cmap="Blues", annot=True)
if not od.empty:
exit()
"""
Example: Comparing multiple weather stations
Meteorological data provided by Meteostat (https://dev.meteostat.net)
under the terms of the Creative Commons Attribution-NonCommercial
4.0 International Public License.
The code is licensed under the MIT license.
"""
from datetime import datetime
import matplotlib.pyplot as plt
from meteostat import Daily
# Time period
start = datetime(2019, 1, 1)
end = datetime(2019, 12, 31)
# Get daily data
data = Daily(['71624', '72295', '68816', '94767'], start, end)
data = data.fetch()
# Plot chart
data.unstack('station')['tavg'].plot(
legend=True,
ylabel='Avg. Daily Temperature °C',
title='Average Temperature Report for 2019')
plt.show()
def stations_daily():
"""
Return daily station data in JSON format
"""
# Get query parameters
args = utils.get_parameters(parameters)
# Check if required parameters are set
if args['station'] and len(args['start']) == 10 and len(args['end']) == 10:
try:
# Convert start & end date strings to datetime
start = datetime.strptime(args['start'], '%Y-%m-%d')
end = datetime.strptime(f'{args["end"]} 23:59:59',
'%Y-%m-%d %H:%M:%S')
# Get number of days between start and end date
date_diff = (end - start).days
# Check date range
if date_diff < 0 or date_diff > max_days:
# Bad request
abort(400)
# Caching
now_diff = (datetime.now() - end).days
if now_diff < 30:
cache_time = 60 * 60 * 24
else:
cache_time = 60 * 60 * 24 * 3
Daily.max_age = cache_time
# Get data
data = Daily(args['station'], start, end, model=args['model'])
# Check if any data
if data.count() > 0:
# Normalize data
data = data.normalize()
# Aggregate
if args['freq']:
data = data.aggregate(args['freq'])
# Unit conversion
if args['units'] == 'imperial':
data = data.convert(units.imperial)
elif args['units'] == 'scientific':
data = data.convert(units.scientific)
# Fetch DataFrame
data = data.fetch()
# Convert to integer
data['tsun'] = data['tsun'].astype('Int64')
# DateTime Index to String
data.index = data.index.strftime('%Y-%m-%d')
data.index.rename('date', inplace=True)
data = data.reset_index().to_json(orient="records")
else:
# No data
data = '[]'
# Inject meta data
meta = {}
meta['generated'] = datetime.now().strftime('%Y-%m-%d %H:%M:%S')
# Generate output string
output = f'''{{"meta":{json.dumps(meta)},"data":{data}}}'''
# Return
return utils.send_response(output, cache_time)
except BaseException:
# Bad request
abort(400)
else:
# Bad request
abort(400)
def get_raw_data(station_id, start=start, end=end):
data_daily = Daily([station_id], start=start, end=end)
return data_daily.fetch()