def get_repetition_times_knmi(dates, pref=None):
"""Get repetition times from KNMI data.
Calculates the repetition times for maximum peak wind and
highest hourly wind peak.
Parameters
----------
dates : list of dt.DateTime objects
Dates of the storm
pref : str or None (optional)
Preference for a certain function.
If None, returns optimal fit function.
If 'gumbel', returns only the Gumbel one.
If 'weibull', returns only the Weibull result.
Returns
-------
rtdf : pandas.DataFrame
DataFrame with location and both types of return times.
"""
rtdf = pd.DataFrame(
columns=['Station_', 'longitude', 'latitude', 'RT', 'RTH'])
for station in knmi.stations.keys():
if station > 199:
data = knmi.get_day_data_dataframe(stations=[station],
start=dt.datetime(1963, 1, 1))
if len(data['FHX']) < 100:
continue
data['FHX'] = data['FHX'] * 0.1
data['FXX'] = data['FXX'] * 0.1
try:
ecdfw = ECDF(data['FXX'][data['FXX'] >= 0])
ecdfh = ECDF(data['FHX'][data['FHX'] >= 0])
lastmax = np.nanmax([data['FXX'][date] for date in dates])
lastmaxh = np.nanmax([data['FHX'][date] for date in dates])
chance = optimal_function(ecdfw.x[1:],
ecdfw.y[1:],
lastmax,
pref=pref)
chanceh = optimal_function(ecdfh.x[1:],
ecdfh.y[1:],
lastmaxh,
pref=pref)
except:
continue
returntime = round(1 / (1 - chance))
returntimeh = round(1 / (1 - chanceh))
if returntime > 0 and returntime < 1000000 and len(
data['FXX'][data['FXX'] >= 0]) > 10 * 365:
if returntimeh > 0 and returntimeh < 1000000 and len(
data['FHX'][data['FHX'] >= 0]) > 10 * 365:
rtdf = rtdf.append(
{
'Station_': int(station),
'longitude': knmi.stations[station].longitude,
'latitude': knmi.stations[station].latitude,
'RT': returntime,
'RTH': returntimeh
},
ignore_index=True)
rtdf['Station_'] = rtdf['Station_'].astype(int)
rtdf['RT'] = rtdf['RT'].astype(int)
rtdf['RTH'] = rtdf['RTH'].astype(int)
return rtdf
# encoding: utf-8 """ Going to use some knwi weather stuff here.. https://github.com/EnergieID/KNMI-py """ import knmi df = knmi.get_day_data_dataframe(stations=[260]) print(df.disclaimer) print(df.stations)
import knmi
df = knmi.get_day_data_dataframe(
stations=[260],
start="19570701",
end="20190701",
inseason=False,
variables=["RH", "EV24"],
)
df.loc[df["RH"] < 0, "RH"] = 0.5
df["RH"] /= 10
df["EV24"] /= 10
df["NOV"] = df["RH"] - df["EV24"]
df["hydr_jaar"] = df.index.year
df.loc[df.index.month < 4, "hydr_jaar"] -= 1
df.groupby("hydr_jaar")["NOV"].sum().to_csv("data/neerslagoverschot.csv", header=True)
#https://github.com/EnergieID/KNMI-py
import knmi
import metadata
print(metadata.stations) # provides a list of all available stations
print(metadata.variables) # provides a list of all variables and explanations
station_de_bilt = 260
station_eelde = 280
print("Ophalen dagdata")
disclaimer, stations, legend, df = knmi.get_day_data_dataframe(
stations=[station_de_bilt, station_eelde],
start="20200501",
end="20200601")
print(stations, legend)
print(df)
print("Ophalen uurdata")
disclaimer, stations, legend, df = knmi.get_hour_data_dataframe(
stations=[station_de_bilt, station_eelde],
start="2020080101",
end="2020080105")
print(stations, legend)
print(df)
print("Ophalen uurdata alleen temp en zon")
disclaimer, stations, legend, df = knmi.get_hour_data_dataframe(
stations=[station_de_bilt, station_eelde],
start="2020080101",
end="2020080124",
variables=['TEMP', 'SUNR'])
print(stations, legend)