def plot_scatter(self,
keys=None,
colwrap=3,
sharex='none',
s=0,
mask='RV_mask',
aggr=None,
title=None):
import df_ana
df_d = self.df_data.mean(axis=0, level=1)
if mask is None:
tv = self.df_data.loc[0].iloc[:, 0]
df_d = df_d
elif mask == 'RV_mask':
tv = self.df_data.loc[0].iloc[:, 0][self.RV_mask.loc[s]]
df_d = df_d[self.RV_mask.loc[s]]
else:
tv = self.df_data.loc[0].iloc[:, 0][mask]
df_d = df_d[mask]
kwrgs = {'tv': tv, 'aggr': aggr, 'title': title}
df_ana.loop_df(df_d,
df_ana.plot_scatter,
keys=keys,
colwrap=colwrap,
sharex=sharex,
kwrgs=kwrgs)
return
def plot_scatter(cls,
keys=None,
colwrap=3,
sharex='none',
s=0,
mask='RV_mask',
aggr=None,
title=None):
df_d = cls.df_data.loc[s]
if mask is None:
tv = cls.df_data.loc[0].iloc[:, 0]
df_d = df_d
elif mask == 'RV_mask':
tv = cls.df_data.loc[0].iloc[:, 0][cls.RV_mask.loc[s]]
df_d = df_d[fcev.RV_mask.loc[s]]
else:
tv = cls.df_data.loc[0].iloc[:, 0][mask]
df_d = df_d[mask]
kwrgs = {'tv': tv, 'aggr': aggr, 'title': title}
df_ana.loop_df(df_d,
df_ana.plot_scatter,
keys=keys,
colwrap=colwrap,
sharex=sharex,
kwrgs=kwrgs)
return
def apply_df_ana_plot(self, df=None, func=None,
hspace=.4, sharex=False, sharey=False,
kwrgs_func={}):
if df is None:
df = self.df_data
if func is None:
func = df_ana.plot_ac ; kwrgs_func = {'AUC_cutoff':(14,30),'s':60}
return df_ana.loop_df(df, function=func, sharex=sharex, sharey=sharey,
colwrap=2, hspace=hspace, kwrgs=kwrgs_func)
q=q_sp,
tailmean=True,
selbox=selbox)
q_sp = 90
ds[f'q{q_sp}tail'] = cl.percentile_cluster(var_filename,
xrclust,
q=q_sp,
tailmean=True,
selbox=selbox)
df_clust = functions_pp.xrts_to_df(ds['ts'])
fig = df_ana.loop_df(df_clust,
function=df_ana.plot_ac,
sharex=False,
colwrap=2,
kwrgs={
'AUC_cutoff': (14, 30),
's': 60
})
fig.suptitle('q: {}, n_clusters: {}'.format(q, c), x=.5, y=.97)
df_clust = functions_pp.xrts_to_df(ds[f'q{q}tail'])
fig = df_ana.loop_df(df_clust,
function=df_ana.plot_ac,
sharex=False,
colwrap=2,
kwrgs={
'AUC_cutoff': (14, 30),
's': 60
})
import df_ana, functions_pp, validation
import pandas as pd
f_format = '.pdf'
fc = list_of_fc[0]
# if os.path.isdir(fc.filename) == False : os.makedirs(fc.filename)
# filepath = '/'.join(fc.filename.split('/')[:-2])
filepath = '/Users/semvijverberg/surfdrive/MckinRepl/ERA5_mx2t_sst_Northern/c378f_ran_strat10_s30/figures'
df_daily = fc.df_data_orig.loc[0][['mx2t'
]].rename(columns={'mx2t': 'ERA5 mx2t'})
df_15 = fc.TV.fullts.rename(columns={'mx2t': 'ERA5 mx2t 15-day mean'})
print(validation.get_bstrap_size(df_daily))
fig = df_ana.loop_df(df_daily,
function=df_ana.plot_ac,
colwrap=1,
kwrgs={'AUC_cutoff': False})
fig.savefig(filepath + '/ac_daily' + f_format, bbox_inches='tight')
fig = df_ana.loop_df(df_15,
function=df_ana.plot_ac,
colwrap=1,
kwrgs={'AUC_cutoff': False})
fig.savefig(filepath + '/ac_15-daymean' + f_format, bbox_inches='tight')
EC_data = user_dir + '/surfdrive/output_RGCPD/easternUS_EC/EC_tas_tos_Northern/958dd_ran_strat10_s30/data/EC_21-03-20_16hr_lag_10_958dd.h5'
fc = fcev(path_data=EC_data,
precur_aggr=precur_aggr,
use_fold=use_fold,
start_end_TVdate=None,
stat_model=None,
fc = list_of_fc[-1]
df, fig = fc.plot_feature_importances(lag=lag_rel)
path_feat = filename + f'ifc{1}_logitregul_l{lag_rel}' + f_format
fig.savefig(path_feat, bbox_inches='tight')
import df_ana
# X_CPPA = list_of_fc[-1]._get_statmodelobject().X_pred[['0..4..sst', '0..6..sst', '0..CPPAsv']]
# X_sm = list_of_fc[-2]._get_statmodelobject().X_pred[['0..2..sm2', '0..2..sm3']]
# X_merge = X_sm.merge(X_CPPA, left_index=True, right_index=True)
# All = fc.TV.RV_ts.merge(X_merge, left_index=True, right_index=True)
All = fc.df_data.loc[0][[
'1q65tail', '0..2..sm2', '0..2..sm3', '0..4..sst', '0..6..sst', '0..CPPAsv'
]]
df_ana.loop_df(df=All, function=df_ana.plot_ac, kwrgs={'s': 150})
fc.dict_sum[0].loc['Precision'].loc['Precision']
fc.dict_sum[0].loc['Accuracy'].loc['Accuracy']
#%%
# im = 0
# il = 1
# ifc = 1
# f_format = '.pdf'
# if os.path.isdir(fc.filename) == False : os.makedirs(fc.filename)
# import valid_plots as dfplots
# if __name__ == "__main__":
# for ifc, fc in enumerate(list_of_fc):
