def get_data(symbol, field, start_date, end_date):
data = ice.get_timeseries(symbol, field, granularity = 'D', start_date= start_date, end_date = end_date)
return data
def get_climo():
field = 'GWDD official 30yr Climo'
start_date_climo = (datetime(2022, 1, 1)).strftime('%Y-%m-%d')
end_date_climo = (datetime(2022, 12, 31)).strftime('%Y-%m-%d')
symbol = location + '-CLMO'
climo = get_data([symbol], field, start_date_climo, end_date_climo)
return climo
def get_history(climo_type, start_date_history, end_date_history):
field = 'GWDD'
symbol = [location + '-' + climo_type]
history = get_data(symbol, field, start_date_history, end_date_history)
return history
def get_fcst(model_run_start_date, model_run_end_date):
fcst_series1=[]; fcst_series2=[];fcst_series3=[];fcst_series4=[];fcst_series5=[]
for model_run_date in np.arange(model_run_start_date, model_run_end_date, timedelta(days=1)).astype(datetime):
model_run_date_str = model_run_date.strftime('%y%m%d')
symbol = location + ' MR' + model_run_date_str + '_' + model_run_and_name
start_date = model_run_date.strftime('%Y-%m-%d')
end_date = (model_run_date + timedelta(days=num_of_fcst_days)).strftime('%Y-%m-%d')
field = ['GWDD', 'GWDD CHNG 1 Forecast Prior', 'GWDD CHNG 2 Forecast Prior', 'GWDD CHNG 4 Forecast Prior','GWDD official 30yr Anomaly']
fcst = get_data(symbol, field, start_date, end_date)
df_data = pd.DataFrame(list(fcst))
df_data = df_data.iloc[1:]
df_data = df_data.iloc[:,1:]
fcst_total = round(df_data,1)
fcst_series1 = np.append(fcst_series1, fcst_total[1])
fcst_series2 = np.append(fcst_series2, fcst_total[2])
fcst_series3 = np.append(fcst_series3, fcst_total[3])
fcst_series4 = np.append(fcst_series4, fcst_total[4])
fcst_series5 = np.append(fcst_series5, fcst_total[5])
return fcst_series1, fcst_series2, fcst_series3, fcst_series4, fcst_series5
def get_df(data, data_name):
df = pd.DataFrame(list(data))
df_history = df.iloc[1:]
df_history = df_history.rename(columns={0:'Date', 1:'GWDD'})
df_history['Date'] = pd.DatetimeIndex(df_history['Date'])
df_history['GWDD'] = pd.to_numeric(df_history['GWDD'])
df_history = df_history.set_index(['Date'])
df_history = df_history.drop(df_history[(df_history.index.month == 2) & (df_history.index.day == 29)].index)
return df_history
def get_df_fcst(fcst1, fcst2, fcst3, fcst4, fcst5, index):
df_data = pd.DataFrame({'GWDD':fcst1,'Change_1':fcst2,'Change_2':fcst3,'Change_4':fcst4,'Anomaly':fcst5, 'Date':index})
date_rng = pd.DatetimeIndex(df_data['Date'])
df_data['Month'] = date_rng.month.values
df_data['Day'] = date_rng.day.values
return df_data
def get_df_anomaly(df_history, df_climo):
df_merge = pd.merge(df_history, df_climo, on=['Month', 'Day'], how='inner') # or outer for the entire year
df_merge = df_merge[['Date_x', 'History', 'Climo']]
df_merge = df_merge.rename(columns={'Date_x':'Date'})
df_merge['Anomaly'] = df_merge['History'] - df_merge['Climo']
df_merge = df_merge.sort_values(by='Date')
df_anomaly = df_merge[['Date', 'Anomaly']]
df_anomaly = df_anomaly.set_index('Date')
df_anomaly.index = pd.to_datetime(df_anomaly.index).date
return df_anomaly
def merge_CFSR_and_METAR(df_history_METAR, df_history_CFSR):
df_merge = pd.merge(df_history_METAR, df_history_CFSR, how='outer', left_index=True, right_index=True)
df_merge['GWDD_y'] = df_merge['GWDD_y'].fillna(df_merge['GWDD_x'])
df_merged_CFSR_and_METAR = df_merge.drop('GWDD_x', axis=1)
df_merged_CFSR_and_METAR = df_merged_CFSR_and_METAR.rename(columns={'GWDD_y':'GWDD'})
date_rng = df_merged_CFSR_and_METAR.index
df_merged_CFSR_and_METAR['Date'] = date_rng
df_merged_CFSR_and_METAR['Month'] = date_rng.month.values
df_merged_CFSR_and_METAR['Day'] = date_rng.day.values
df_merged_CFSR_and_METAR = df_merged_CFSR_and_METAR.rename(columns={'GWDD':'History'})
return df_merged_CFSR_and_METAR
def plot_time_series(df_history_and_fcst_anomaly):
fig, ax1 = plt.subplots(1,1, figsize=(16,9), dpi=100)
vertical_line = df_history_and_fcst_anomaly.index[df_history_and_fcst_anomaly.index==model_run_start_date.date()]
tseries = df_history_and_fcst_anomaly['Anomaly']
color = (tseries > 0).apply(lambda x: 'green' if x else 'brown')
ax1.bar(df_history_and_fcst_anomaly.index, df_history_and_fcst_anomaly['Anomaly'], color=color, label = 'GWDD Anomaly')
ax1.set_ylabel('GWDD Anomaly',fontsize=18)
ax1.set_ylim(-20, 20)
ax1.set_yticks(np.linspace(ax1.get_yticks()[0], ax1.get_yticks()[-1], len(ax1.get_yticks())))
ax1.axhline(0, color='grey', linewidth=0.8)
ax1.xaxis.set_major_locator(mdates.DayLocator(interval=2))
ax1.xaxis.set_major_formatter(mdates.DateFormatter('%m-%d-%Y'))
plt.gcf().autofmt_xdate(rotation=45)
ax1.legend()
ax1.set_title('Daily ' + location + ' GWDD Anomalies (Relative to the 30-yr Normal)\n'
+ 'Constructed from: Observations and ' + model_run_and_name + ' Forecast\n'
+ 'Model Run Date: ' + now.strftime('%a, %b %d'), fontsize=18)
ax1.grid()
ax1.axvline(vertical_line[0]-timedelta(hours=12),color='black', linewidth=2)
plt.tight_layout()
plt.show()
