综合实训 长沙气候数据分析

第一题

for line in f1:
    nf = line[11:15]  
    yf = line[15:17]  
    lx= line[17:21]  
    if "s" in line:
        sj=line[22:].strip().split("s")
    else:
        sj=line[22:].strip().split("S")
    for i in range(0, len(sj)):
        if i+1<10:
            rq = nf+"-"+yf + '-0' + str(i+1)  
        else:
            rq = nf+"-"+yf+ '-' + str(i+1) 
        sj[i]=sj[i].strip().replace("T","")
        if   len(sj[i])>0 and  not ("-9999" in sj[i]):
            js= float(sj[i])/10
            if lx=="TMAX" :
                zd[rq]=[js]
            else:
                zd[rq].append(js)

第二题

dfmr = pd.DataFrame(list(zd.values()),index=zd.keys(),columns=["gw","dw","sl","pj"]) 
pjjs = dfmr["sl"].mean()
pjwd = dfmr["pj"].mean()
zgwd = dfmr['gw'].max()
zdwd = dfmr['dw'].min()
dfmr[dfmr['gw']==zgwd].index
zgwdrq=list(dfmr[dfmr['gw']==zgwd].index)
zdwdrq=list(dfmr[dfmr['dw']==zdwd].index)

第三题 

dfmr = pd.read_csv('data/zd.txt', index_col=0,header=None,names=["gw","dw","sl","wd"])  
df_yftj=dfmr.groupby(dfmr.index.str[5:7]).agg({'sl':'mean','wd':'mean',"gw":"mean","dw":"mean"})

第四题 

 

plt.figure(figsize=(20,10))  
sj=range(1,13,1)  
plt.bar(sj,df_yftj["sl"])  
for i in range(12):  
    plt.text(i+0.9,df_yftj["sl"][i]+0.2,"{:.2f}".format(df_yftj["sl"][i]))  
plt.xlabel("月")  
plt.ylabel("雨量")  
plt.title("各月日均降水量")  
plt.xticks(sj)  
plt.ylim(0,8)  
plt.grid(True) 

第五题 

sj=range(1,13)
plt.figure(figsize=(20,10))
plt.plot(sj,df_yftj["wd"],label="平均温度")
plt.plot(sj,df_yftj["gw"],"r",label="最高温度")
plt.plot(sj,df_yftj["dw"],"g",label="最低温度")
for i in range(12):
    plt.text(i+0.8,df_yftj["wd"][i]+1,"{:.2f}".format(df_yftj["wd"][i]))
plt.xlabel("月")
plt.ylabel("气温")
plt.title("各月日均温度")
plt.xticks(sj)
plt.grid()
plt.xlim(1,12)
plt.legend(loc="upper right")

第六题 

 

wd=[]
f2=open("data2/wdyc.txt","w")
for line in f1:
    lx= line[17:21]  
    if lx=="TAVG" :
        if "s" in line:
            sj=line[22:].strip().split("s")
        else:
            sj=line[22:].strip().split("S")
        for i in range(0, len(sj)):
            sj[i]=sj[i].strip().replace("T","")
            if   len(sj[i])>0 and  not ("-9999" in sj[i]):
                js= float(sj[i])/10
                wd.append(str(js))
                if len(wd)==11 :
                    f2.write(",".join(wd)+"\n")
                    wd.pop(0)
f2.close()

第七题

import pandas as pd
data=pd.read_csv("data2//wdyc.txt",header=None)
x=data.iloc[:,0:10]
y=data.iloc[:,10]
from sklearn.cross_validation import train_test_split
x_train,x_test,y_train,y_test=train_test_split(x,y,random_state=1,train_size=0.8)
from sklearn.svm import SVR
clf = SVR(kernel='rbf',C=6,gamma=0.01)
clf.fit(x_train,y_train)
y_predict=clf.predict(x_test)
wc=0
for i in range(len(x_test)):    
    wc=wc+abs(y_predict[i]-y_test.iloc[i])
print("平均误差{:.2f}".format(wc/len(x_test)))