-
Notifications
You must be signed in to change notification settings - Fork 0
/
precip_functions.py
271 lines (248 loc) · 12.4 KB
/
precip_functions.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
#=================================
# This Python defines different functions for analyzing the precipitation data
# Edited by Yi Hong, Oct. - Nov. 2019
#=================================
#%% import packages
import numpy as np
import pandas as pd
#import xarray as xr
import datetime
import os
import re
from math import cos, asin, sqrt
#%% =============================================================================
# Define a function to read the rain gauge data from generated files
# Return a list with the required data
# =============================================================================
def Read_rain_file(record_dic, gauge_file, begin_date, end_date, date_format):
# dir_ReCON=record_dic+'ReCON_stations\\'
begin_time=datetime.datetime.strptime(begin_date, date_format)
end_time=datetime.datetime.strptime(end_date,date_format)
f=open(os.path.join(record_dic, gauge_file), 'r', errors=None)
record_file=f.readlines()
f.close()
begin_line=record_file[0].split(';')
record_begin=datetime.datetime.strptime(str(begin_line[0]),date_format)
end_line=record_file[-1].split(';')
record_end=datetime.datetime.strptime(str(end_line[0]), date_format)
if begin_time<(record_begin-datetime.timedelta(days=366)): # have to consider if record begins in the middle of a year
print('Please reset a begin date')
return []
elif end_time>(record_end+datetime.timedelta(days=366)):
print('Please reset an end date') # have to consider if record ends in the middle of a year
return []
else:
i=0
j=len(record_file)-1
while datetime.datetime.strptime(str(record_file[i].split(';')[0]),date_format)<begin_time:
i+=1
while datetime.datetime.strptime(str(record_file[j].split(';')[0]),date_format)>end_time+ datetime.timedelta(days=1):
j-=1
return record_file[i:j]
# test=[s for s in record_file if '2015/12/31' in s] ## cannot use this, cause the data is not complete
#%%
# =============================================================================
# Resample the precip record with the specified frequency of input and output data
# Set the limit for nan data
# =============================================================================
def resamp_precip(freq_in, freq_out, resample_file, min_count, begin_date, end_date, date_format):
#% it is better to use the pandas dataframe for the calculations
df = pd.read_csv(resample_file,delimiter=';',names=['Time','Precip'])
df['DateTime'] = pd.to_datetime(df['Time'],format=date_format)
df=df.drop(['Time'],axis=1)
#% Use a date time index from the beginning to the end at the specified time step
# record_begin=df['DateTime'].iloc[0]
# record_end=df['DateTime'].iloc[-1]
begin_time=datetime.datetime.strptime(begin_date,date_format)
end_time=datetime.datetime.strptime(end_date,date_format)
#(end_time-begin_time)/datetime.timedelta(minutes=30)
#% Not necessary to test the begin time
# if begin_time<record_begin or end_time>record_end: # have to consider if record begins in the middle of a year
# print('Please reset the dates')
# return []
# else:
#% Slice the data and set datetime index
df_query=df.loc[(df['DateTime']>=begin_time) & (df['DateTime']<=end_time)].drop_duplicates() #df.between_time
# df_query[df_query.index.duplicated()]
df_query=df_query.set_index('DateTime')
df_query=df_query.sort_index().to_period(freq_in)
#print('shape of the analyzed dataframe: ',df_query.shape)
#df_query.plot(y='Precip')
#% Fill in with the full 30 min data
rng_in = pd.period_range(begin_time,end_time,freq=freq_in)
df_query=df_query.reindex(rng_in)
na_in=df_query['Precip'].isna().sum()
#df_query.asfreq(freq=freq_in, fill_value=-999)
#print('Test of the precip sum: ',df_query['Precip'].sum())
#print('Nans in the new precip serie: ',na_in)
#%
precip_out=df_query.resample(freq_out).sum(min_count=min_count) # sum(skipna=true)
na_out=precip_out['Precip'].isna().sum()
#%
return precip_out, na_in, na_out
#%% Resample a dataframe with specified frequency of input and output data
### the df is already pre-treated with a datetime index and sort
def resamp_df(df, freq_in, freq_out, min_count, begin_date, end_date, date_format):
begin_time=datetime.datetime.strptime(begin_date,date_format)
end_time=datetime.datetime.strptime(end_date,date_format)
df_query=df.loc[(df.index>=begin_time) & (df.index<=end_time)] #df.between_time
df_query=df_query.to_period(freq_in)
rng_in = pd.period_range(begin_time,end_time,freq=freq_in)
df_query=df_query.reindex(rng_in)
na_in=df_query.iloc[:,0].isna().sum()
precip_out=df_query.resample(freq_out).sum(min_count=min_count) # sum(skipna=true)
na_out=precip_out.iloc[:,0].isna().sum()
return precip_out, na_in, na_out
#%%
# =============================================================================
# # fonction to select AORC data correspond to rain gauge stations
# =============================================================================
def AORC_raingauge(ds,index):
if ds['RAINRATE'].size>0 and ds['valid_time'].size>0:
rain_array=ds['RAINRATE']
rain_intense=rain_array[dict(Time=0, south_north=index[0], west_east=index[1])].data # mm s^-1
rain_time=ds['valid_time'].data
return rain_time,rain_intense
#%%
# =============================================================================
# Function to extract the date from the filename by using the regular expression to find
# =============================================================================
def date_filename(filename):
date_format='20[01]\d[01]\d[0123]\d.nc' # date format
# date_format=set_year(year)
date_nc=re.findall(date_format,filename)
if len(date_nc)>0:
str_date=date_nc[0][0:len(date_nc[0])-3]
return str_date
else:
return []
# =============================================================================
# Functions to calculate the RMSE and MAE, and (prod-obs)/obs
# =============================================================================
def MAE_RMSE_Diff(serie1, serie2):
if len(serie1)==len(serie2):
serie1=serie1.reset_index(drop=True)
serie2=serie2.reset_index(drop=True)
MAE=(serie1-serie2).abs().mean()
RMSE=np.sqrt(((serie1-serie2)**2).mean())
Diff=(serie1-serie2).sum(min_count=1)
return MAE, RMSE, Diff
else:
raise Exception('Length of the two series are not equal')
return []
def PBIAS(serie1, serie2): # this function is to calculate the Diff/Obs, the problem is when obs is low, this value may over estimate
if len(serie1)==len(serie2):
if serie1.isnull().all() or serie2.isnull().all():
return np.nan
else:
Total_dif=0
Total_obs=0
for i in range(len(serie1)):
if (not np.isnan(serie1.iloc[i]) and not np.isnan(serie2.iloc[i])):
Total_dif += serie1.iloc[i]-serie2.iloc[i]
Total_obs += serie2.iloc[i]
if Total_obs>0:
PBIAS= 100 * Total_dif/Total_obs
return PBIAS
else: # if Total_obs ==0, how to set the diff ?
# return serie1.sum()
return np.nan
else:
raise Exception('Length of the two series are not equal')
return np.nan
def R2(serie1, serie2): # this function is to calculate the determinant R2
if len(serie1)==len(serie2):
if serie1.isnull().all() or serie2.isnull().all():
return np.nan
else: # Compute only when both S1 and S2 is true
S1=[]
S2=[]
for i in range(len(serie1)):
if (not np.isnan(serie1.iloc[i]) and not np.isnan(serie2.iloc[i])):
S1.append(serie1.iloc[i])
S2.append(serie2.iloc[i])
if len(S1)==0:
return np.nan
else:
mean_S1=sum(S1)/len(S1)
mean_S2=sum(S2)/len(S2)
diff_S12=0
diff2_S1=0
diff2_S2=0
for j in range(len(S1)):
diff_S12 += (S1[j]-mean_S1)*(S2[j]-mean_S2)
diff2_S1 += (S1[j]-mean_S1)**2
diff2_S2 += (S2[j]-mean_S2)**2
if diff2_S1 != 0 and diff2_S2 != 0:
R2 = diff_S12/(np.sqrt(diff2_S1)*np.sqrt(diff2_S2))
return R2
else:
return np.nan
else:
raise Exception('Length of the two series are not equal')
return np.nan
# =============================================================================
# Function to plot the base map
# =============================================================================
def map_area(shorline_file):
sl_coor= pd.read_csv(shorline_file, sep=r'\s+', skiprows=2, header=None, index_col=False, skipinitialspace=True,engine='python', names=['Lon','Lat'])
min_Lon=min(sl_coor['Lon'])
max_Lon=max(sl_coor['Lon'].loc[sl_coor['Lon']<0])
min_Lat=min(sl_coor['Lat'].loc[sl_coor['Lat']>40])
max_Lat=max(sl_coor['Lat'].loc[sl_coor['Lat']<50])
return min_Lon,max_Lon,min_Lat,max_Lat
#%
#%%
# =============================================================================
# Function to calculat the distance between two points with given lat and lon
# =============================================================================
# The Haversine formula is needed for a correct calculation of the distance between points on the globe
def distance(lat1, lon1, lat2, lon2):
p = 0.017453292519943295 # Math.PI / 180
a = 0.5 - cos((lat2-lat1)*p)/2 + cos(lat1*p)*cos(lat2*p) * (1-cos((lon2-lon1)*p)) / 2
return 12742 * asin(sqrt(a)) #2 * R; R = 6371 km
#%%
def closest_index(list_coord, point_coord):
min_dist=min(list_coord, key=lambda point: distance(point_coord['Lat'],point_coord['Lon'],point['Lat'],point['Lon']))
return list_coord.index(min_dist)
#
#def closest_xy(list_coord, point_coord):
# min_dist=min(list_coord, key=lambda point: distance(point_coord['Lat'],point_coord['Lon'],point['Lat'],point['Lon']))
# return list_coord.index(min_dist)
#%%
def read_hrrr_var(hrrr_ds, var_name, y, x, date_format):
ds_precip=hrrr_ds[var_name].isel(y=y,x=x)
ds_precip=ds_precip.resample(time='1D').sum("time").reset_coords(drop=True)
date_list=pd.to_datetime(ds_precip.time.values).strftime(date_format).values.tolist()
if var_name=='PRATE_surface':
precp_array=3600*ds_precip.values.astype('float')
else:
precp_array=ds_precip.values.astype('float')
precp_array[precp_array < 0] =np.nan
precip_list=precp_array.tolist()
return date_list, precip_list
#%%
#list_coord = [{'lat': 39.7612992, 'lon': -86.1519681},
# {'lat': 39.762241, 'lon': -86.158436 },
# {'lat': 39.7622292, 'lon': -86.1578917}]
#
#point_coord = {'lat': 39.7622290, 'lon': -86.1519750}
#print(closest(tempDataList, v))
##%%
#import mputil
#def distance2(point1, point2):
# return mputil.haversine_distance(point1, point2)
#%%
# =============================================================================
# Fonction to filter the gauge stations, input is the serie of na analysis with stn name as index
# =============================================================================
def stn_filter(ivar,stn_info,limit_na,df_anal):
#%
stn_out=pd.DataFrame(columns=stn_info.columns)
df_anal_out=pd.DataFrame(index=df_anal.index)
for istn in df_anal.columns:
if df_anal.loc[ivar,istn]<=limit_na:
stn_out=stn_out.append(stn_info.loc[stn_info['Stn_id']==istn])
df_anal_out=df_anal_out.join(df_anal.loc[:,istn])
#%
return stn_out, df_anal_out