-
Notifications
You must be signed in to change notification settings - Fork 0
/
5002_data_analysis.py
254 lines (197 loc) · 11.1 KB
/
5002_data_analysis.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
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Fri Nov 23 19:15:45 2018
@author: junewang
"""
import pandas as pd
import matplotlib.pyplot as plt
import time
import numpy as np
import seaborn as sns
from statsmodels.tsa.arima_model import ARIMA
from statsmodels.graphics.tsaplots import plot_acf, plot_pacf
import statsmodels.api as sm
station_label=pd.read_csv("station_label.csv",header=None)
aiqQuality_201804=pd.read_csv("aiqQuality_201804.csv")
airQuality_201701_201801=pd.read_csv("airQuality_201701-201801.csv")
airQuality_201802_201803=pd.read_csv("airQuality_201802-201803.csv")
Beijing_grid_weather_station=pd.read_csv("Beijing_grid_weather_station.csv",header=None)
Beijing_grid_weather_station.columns = ['station_id', 'longitude', 'latitude']
gridWeather_201701_201803=pd.read_csv("gridWeather_201701-201803.csv")
gridWeather_201804=pd.read_csv("gridWeather_201804.csv")
gridWeather_20180501_20180502=pd.read_csv("gridWeather_20180501-20180502.csv")
observedWeather_201701_201801=pd.read_csv("observedWeather_201701-201801.csv")
observedWeather_201802_201803=pd.read_csv("observedWeather_201802-201803.csv")
observedWeather_201804=pd.read_csv("observedWeather_201804.csv")
observedWeather_20180501_20180502=pd.read_csv("observedWeather_20180501-20180502.csv")
aiqQuality_201804=pd.read_csv("aiqQuality_201804.csv")
station_info=pd.read_csv("station.csv",header=None)
station_info.columns = ['station_id', 'longitude', 'latitude']
station=list(station_info["station_id"])
##### 更改列明
station_label.columns = ['station_id', 'longitude', 'latitude','station_type']
del station_label["longitude"]
del station_label["latitude"]
aiqQuality_201804.columns = ['id', 'station_id', 'utc_time','PM2.5','PM10','NO2','CO','O3','SO2']
airQuality_201701_201801.rename(columns={ airQuality_201701_201801.columns[0]: "station_id" }, inplace=True)
airQuality_201802_201803.rename(columns={ airQuality_201802_201803.columns[0]: "station_id" }, inplace=True)
aiqQuality_201804=aiqQuality_201804.drop(['id'], axis=1)
gridWeather_201701_201803.rename(columns={ gridWeather_201701_201803.columns[0]: "station_id" }, inplace=True)
gridWeather_201701_201803.rename(columns={ gridWeather_201701_201803.columns[0]: "station_id" }, inplace=True)
gridWeather_201701_201803.rename(columns={ gridWeather_201701_201803.columns[-1]: "wind_speed" }, inplace=True)
gridWeather_201804.rename(columns={ gridWeather_201804.columns[2]: "utc_time" }, inplace=True)
gridWeather_20180501_20180502.rename(columns={ gridWeather_20180501_20180502.columns[2]: "utc_time" }, inplace=True)
gridWeather_201804=gridWeather_201804.drop(['id'],axis=1)
gridWeather_201701_201803=gridWeather_201701_201803.drop(['longitude','latitude'],axis=1)
gridWeather_20180501_20180502=gridWeather_20180501_20180502.drop(['id'],axis=1)
observedWeather_201804.rename(columns={ observedWeather_201804.columns[2]: "utc_time" }, inplace=True)
observedWeather_20180501_20180502.rename(columns={ observedWeather_20180501_20180502.columns[2]: "utc_time" }, inplace=True)
observedWeather_201701_201801=observedWeather_201701_201801.drop(['longitude','latitude'],axis=1)
observedWeather_201804=observedWeather_201804.drop(['id',],axis=1)
observedWeather_20180501_20180502=observedWeather_20180501_20180502.drop(['id',],axis=1)
concat_airqQuality=pd.concat([aiqQuality_201804,airQuality_201701_201801,airQuality_201802_201803])
concat_gridWeather=pd.concat([gridWeather_201804,gridWeather_201701_201803,gridWeather_20180501_20180502])
concat_observedWeather=pd.concat([observedWeather_201701_201801,observedWeather_201802_201803,observedWeather_201804,observedWeather_20180501_20180502])
concat_airqQuality.sort_values(by=["station_id","utc_time"], inplace=True)
concat_gridWeather.sort_values(by=["station_id","utc_time"], inplace=True)
concat_observedWeather.sort_values(by=["station_id","utc_time"], inplace=True)
concat_airqQuality=concat_airqQuality.drop_duplicates(keep='first')
concat_gridWeather=concat_gridWeather.drop_duplicates(keep='first')
concat_observedWeather=concat_observedWeather.drop_duplicates(keep='first')
concat_airqQuality["time_month"]=concat_airqQuality["utc_time"].apply(lambda x:time.strptime(str(x), "%Y-%m-%d %H:%M:%S")[1])
concat_airqQuality=concat_airqQuality[concat_airqQuality["time_month"].isin([3,4,5,6])]
concat_gridWeather["time_month"]=concat_gridWeather["utc_time"].apply(lambda x:time.strptime(str(x), "%Y-%m-%d %H:%M:%S")[1])
concat_gridWeather=concat_gridWeather[concat_gridWeather["time_month"].isin([3,4,5,6])]
concat_observedWeather["time_month"]=concat_observedWeather["utc_time"].apply(lambda x:time.strptime(str(x), "%Y-%m-%d %H:%M:%S")[1])
concat_observedWeather=concat_observedWeather[concat_observedWeather["time_month"].isin([3,4,5,6])]
unique_station_gridWeather=concat_gridWeather["station_id"].unique()
unique_station_airQuality=list(concat_airqQuality["station_id"].unique())
unique_station_observedWeather=pd.DataFrame(concat_observedWeather["station_id"].unique())
concat_gridWeather=pd.merge(concat_gridWeather,Beijing_grid_weather_station,on="station_id")
concat_airqQuality=pd.merge(concat_airqQuality,station_info,on="station_id")
concat_airqQuality.info()
concat_gridWeather.info()
concat_observedWeather.info()
#concat_airqQuality.to_csv("process_data/concat_airqQuality.csv")
#concat_gridWeather.to_csv("process_data/concat_gridWeather.csv")
#concat_observedWeather.to_csv("process_data/concat_observedWeather.csv")
###plot, the location of the station
plt.scatter(station_info["longitude"],station_info["latitude"])
plt.title('station location distribution')
###missing value
def get_missing_value_number(unique_station_airQuality,concat_airqQuality,col):
missing=pd.DataFrame(columns=["stations","missing_number","missing_ratio"])
for i in range(len(unique_station_airQuality)):
df=concat_airqQuality[concat_airqQuality["station_id"]==unique_station_airQuality[i]]
missing_number=len(df)-df[col].count()
missing_ratio=missing_number/len(df)
dic={"stations":unique_station_airQuality[i],"missing_number":missing_number,"missing_ratio":missing_ratio}
missing=missing.append(dic,ignore_index=True)
return missing
missing=get_missing_value_number(unique_station_airQuality,concat_airqQuality,"PM10")
#concat_airqQuality=pd.read_csv("process/concat_airqQuality.csv")
#concat_gridWeather=pd.read_csv("process/concat_gridWeather.csv")
#concat_observedWeather=pd.read_csv("process/concat_observedWeather.csv")
#concat_airqQuality=concat_airqQuality.iloc[:,1:]
###1. interpolated method1:spatial interpolation
def get_distance_matrix(station_info):
distance_matrix=np.zeros([len(station_info),len(station_info)])
for i in range(len(station_info)):
for j in range(len(station_info)):
dist=pow((station_info.iloc[i,1]-station_info.iloc[j,1]),2)+pow((station_info.iloc[i,2]-station_info.iloc[j,2]),2)
distance_matrix[i,j]=np.sqrt(dist)
return distance_matrix
distance_matrix=get_distance_matrix(station_info)
def get_3_means_station(distance_matrix,unique_station_airQuality):
dic={}
for i in range(len(unique_station_airQuality)):
station_tuple=zip(unique_station_airQuality,list(distance_matrix[:,i]))
station_tuple=sorted(station_tuple,key=lambda t:(-t[1],t[0]))
dic[unique_station_airQuality[i]]=[station_tuple[i] for i in range(3)]
return dic
k_means_dic=get_3_means_station(distance_matrix,unique_station_airQuality)
###空间插值：IDW算法插值，距离越远，占的比重越小
def interpolate_attri(k_means_dic,concat_airqQuality,attri):
def get_weight(k_means_dic):
dicx={}
for k,v in k_means_dic.items():
sum_dist=sum([v[i][1] for i in range(3)])
p=[]
for x in v:
p.append(x[1]/sum_dist)
sum_p=sum(p)
for i in range(len(p)):
p[i]=(v[i][0],round(p[i]/sum_p,3))
dicx[k]=p
return dicx
dicx=get_weight(k_means_dic)
dic_station={}
dic_weight={}
for k,v in dicx.items():
dic_station[k]=[x[0] for x in v]
dic_weight[k]=[x[1] for x in v]
for i in range(len(concat_airqQuality)):
print(i)
if np.isnan(concat_airqQuality.loc[i,attri])==True:
print(i)
df=concat_airqQuality[concat_airqQuality["utc_time"]==concat_airqQuality.loc[i,"utc_time"]]
df=df[df["station_id"].isin(dic_station[concat_airqQuality.loc[i,"station_id"]])]
pre_val=0
df=df.reset_index()
del df['index']
stations=dic_station[concat_airqQuality.loc[i,"station_id"]]
weights=dic_weight[concat_airqQuality.loc[i,"station_id"]]
for j in range(len(df)):
if np.isnan(df.loc[j,attri])==True:
continue
index=stations.index(df.iloc[j,0])
pre_val+=df.loc[j,attri]*weights[index]
if pre_val!=0:
concat_airqQuality.loc[i,attri]=pre_val
return concat_airqQuality
concat_airqQuality=interpolate_attri(k_means_dic,concat_airqQuality,"PM10")
concat_airqQuality=interpolate_attri(k_means_dic,concat_airqQuality,"CO")
concat_airqQuality=interpolate_attri(k_means_dic,concat_airqQuality,"NO2")
concat_airqQuality=interpolate_attri(k_means_dic,concat_airqQuality,"O3")
concat_airqQuality=interpolate_attri(k_means_dic,concat_airqQuality,"PM2.5")
concat_airqQuality=interpolate_attri(k_means_dic,concat_airqQuality,"SO2")
####2. interpolated method2: 利用arima模型预测插值
concat_airqQuality_inpl=concat_airqQuality
stations=list(concat_airqQuality_inpl["station_id"].unique())
def ARMA_MODEL(timeseries): #返回预测值
order=sm.tsa.arma_order_select_ic(timeseries,max_ar=3,max_ma=3,ic='bic')['bic_min_order']
temp_model= sm.tsa.ARMA(timeseries,order).fit()
pred=temp_model.forecast(1) #返回后面一期
return pred[0][-1]
def Arima_interpolate(air_qual,attri,stations):
air_qual_pre=pd.DataFrame()
for station in stations:
print(station)
temp=air_qual[air_qual["station_id"]==station]
temp=temp.reset_index()
del temp['index']
for i in range(len(temp)):
if np.isnan(temp.loc[i,attri])==True:
print(i)
timeseries=temp[attri][:i]
try:
temp.loc[i,attri]=ARMA_MODEL(timeseries)
except:
print("not predit")
air_qual_pre=pd.concat([air_qual_pre,temp],axis=0)
return air_qual_pre
concat_airqQuality_inpl=Arima_interpolate(concat_airqQuality_inpl,"PM2.5",stations)
concat_airqQuality_inpl=Arima_interpolate(concat_airqQuality_inpl,"O3",stations)
concat_airqQuality_inpl=Arima_interpolate(concat_airqQuality_inpl,"SO2",stations)
concat_airqQuality_inpl=Arima_interpolate(concat_airqQuality_inpl,"NO2",stations)
###3. interpolated method3: 时间差值：
def interpolate(df,stations):
res_df=pd.DataFrame()
for station in stations:
dfx=df[df["station_id"]==station]
dfx=dfx.interpolate(method="cubic")
res_df=res_df.append(dfx)
return res_df
concat_airqQuality_inpl=interpolate(concat_airqQuality_inpl,stations)
concat_airqQuality_inpl.to_csv("process_data/concat_airqQuality_all_interpolated.csv")