#coding=utf-8
import numpyas np
import pandasas pd
import os
import re
import datetimeas dt
from matplotlib.datesimport date2num, DateFormatter
import matplotlib.pyplotas plt
import matplotlib.datesas mdates
from matplotlib.font_managerimport FontProperties
font = FontProperties(fname=r"C:\\Windows\\Fonts\\simsun.ttc", size=25)
from mpl_toolkits.axes_grid1import make_axes_locatable
import sys
sys.path.insert(1,r'H:\microwave_radiometer/')
import variables_dictionaryas vd
# vardic=vd.vardic
def create_empty_mother_array_for_dates_and_data(varname):
dates_all=np.array([],dtype=int).reshape(0, 6)#### year,mon,day,hour,min,sec
if vd.vardic[varname]['levels']!=None:
levels=vd.vardic[varname]['levels']
data_all=np.array([]).reshape(0,len(levels))
# print levels
if vd.vardic[varname]['levels']==None:
data_all=np.array([]).reshape(0,1)
return dates_all,data_all
print ""
def draw_2D(varname,times,data_all):
levels=vd.vardic[varname]['levels']
xx,yy=np.meshgrid(times,levels)
# print data_all.shape
fig=plt.figure(figsize=(10,6))
rect=0.1, 0.1, 0.7, 0.8
ax=fig.add_axes(rect)### rect=l, b, w, h
fig.autofmt_xdate()
pcolormesh=ax.pcolormesh(xx,yy,data_all.transpose());
plt.title(unicode(varname,'utf-8')+'(%s)'%unicode(vd.vardic[varname]['unit'],'utf-8'),fontproperties=font)
xfmt = mdates.DateFormatter('%d %H:%M')
ax.xaxis.set_major_formatter(xfmt)
ax.set_xlim(dt.datetime(2020,4,16,20,0,0),dt.datetime(2020,4,17,20,0,0))
#### colorbar
ax_colorbar=fig.add_axes([0.85,0.1,0.05,0.8])
# ax_colorbar=fig.add_axes(rect=[0.85,0.1,0.1,0.8])
fig.colorbar(mappable=pcolormesh,cax=ax_colorbar)
plt.savefig(unicode(varname,'utf-8')+'.png')
plt.close()
def draw_1D(varname,times,data_all,yaxis_range=None):
fig=plt.figure(figsize=(10,6))
rect=0.1, 0.1, 0.7, 0.8
ax=fig.add_axes(rect)
fig.autofmt_xdate()
# print times.shape,data_all.shape
# ax.scatter(times,data_all.transpose().flatten())
ax.plot(times,data_all.transpose().flatten())
#plt.title(unicode(varname,'utf-8')+'(%s)'%unicode(vd.vardic[varname]['unit'],'utf-8'),fontproperties=font)
plt.title(varname,fontproperties=font)
xfmt = mdates.DateFormatter('%d %H:%M')
ax.xaxis.set_major_formatter(xfmt)
ax.set_xlim(dt.datetime(2020,4,16,20,0,0),dt.datetime(2020,4,17,20,0,0))
if yaxis_range==None:pass
if yaxis_range!=None:ax.set_ylim(yaxis_range[0],yaxis_range[-1])
#plt.savefig(unicode(varname,'utf-8')+'.png')
plt.savefig(varname+'.png')
# plt.show()
plt.close()
# dates_list=[]
# data=pd.read_csv(r'H:\microwave_radiometer\2020_04_16_case_study\\20200415_20200417.txt',skiprows=0,header=1,sep=' ')
# dates=np.array(data[['Year','Mon','Day','Hour']],dtype=int)
# rain=np.array(data[['PRE_1h']],dtype=np.float32)
# for year,mon,day,hour in dates:
# # print year,mon,day,hour
# dates_list.append(dt.datetime(year,mon,day,hour)+dt.timedelta(hours=+8))
# print dates_list
# draw_1D(unicode('PRE_1h'),dates_list,rain)
def read_MZ_minute_txt(filepath):
yearmonday_of_file=filepath[-12:-4]
# print yearmondate_of_file
year,mon,day=yearmonday_of_file[0:4],yearmonday_of_file[4:6],yearmonday_of_file[6:8];year,mon,day=map(int,(year,mon,day))
# print year,mon,day
rain_list=[]
time_list=[]
ff=open(filepath,'r')
ff.readline()
for iin range(1440):
line=ff.readline()
rain_list.append(float(line[56:60])/10)
# print dt.datetime(year,mon,day,hour=20,minute=1)+dt.timedelta(minutes=i)+dt.timedelta(days=-1)
time_list.append(dt.datetime(year,mon,day,hour=20,minute=1)+dt.timedelta(minutes=i)+dt.timedelta(days=-1))
return rain_list,time_list
# rain_list1,time_list1=read_MZ_minute_txt(r'E:\microwave_radiometer\2020_04_16_case_study\AWS_M_Z_54237_20200416.txt')
# rain_list2,time_list2=read_MZ_minute_txt(r'E:\microwave_radiometer\2020_04_16_case_study\AWS_M_Z_54237_20200417.txt')
#
# rain_list=rain_list1+rain_list2
# time_list=time_list1+time_list2
# draw_1D(unicode('PRE_1min'),time_list,np.array(rain_list),yaxis_range=[0,1])
def get_data_by_varname_of_micro_and_daterange(varname='水汽廓线',startday='20200501',endday='20200509'):#####
dates_all,data_all=create_empty_mother_array_for_dates_and_data(varname=varname)
year,mon,day=map(int,(startday[:4],startday[4:6],startday[6:]));startday_dt=dt.datetime(year,mon,day)
year,mon,day=map(int,(endday[:4],endday[4:6],endday[6:]));endday_dt=dt.datetime(year,mon,day)
while(startday_dt!=endday_dt+dt.timedelta(days=1)):
result=vd.get_dates_and_data(date=startday_dt.strftime('%Y%m%d'),varname=varname)
if result=="This date has no data":
print "%s has no data of %s"%(startday_dt,varname)
startday_dt=startday_dt+dt.timedelta(days=1)
continue
print startday_dt
dates_array,data_array=result
dates_all=np.concatenate([dates_all,dates_array])
data_all=np.concatenate([data_all,data_array])
startday_dt=startday_dt+dt.timedelta(days=1)
return dates_all,data_all##### dates_all[0]=(year,mon,day,hour,min,sec)
# dates_all,data_all=get_data_by_varname_of_micro_and_daterange(varname='水汽廓线',startday='20200501',endday='20200510')
# print dates_all
# exit()
hourlyrainfaile=open(r'H:\microwave_radiometer\something_about_data_usage\2019june-2020june_hourly_rain.txt')
dates_rain=[];data_rain=[]
hourlyrainfaile.readline()
while True:
line=hourlyrainfaile.readline()
if line=='':
break
ss=line.split(',')
# station=ss[5];year=ss[]
year=ss[10];mon=ss[11];day=ss[12];hour=ss[13];rain_1h=ss[-2]
year,mon,day,hour=map(int,(year,mon,day,hour));rain_1h=float(rain_1h)
# print ss
# print "fasdfasdfasdf"
dates_rain.append(dt.datetime(year,mon,day,hour))
data_rain.append(rain_1h)
data_rain=np.array(data_rain)
print data_rain.shape
print np.corrcoef(data_rain,data_rain)
exit()
font = FontProperties(fname=r"C:\\Windows\\Fonts\\simsun.ttc", size=45)
def draw_vardata_1D_and_rain_2yaxis(varname,vardata,vardata_times,raindata,rain_times,xaxis=['201907110000','202006100000']):#xaxis=['202004162130','202004162300']
print vardata.shape,len(raindata)
# print raindata
fig, ax1 = plt.subplots(figsize=(15,9),tight_layout=True)
ax2=ax1.twinx()
# a1=ax1.scatter(rain_times, raindata,s=1, c='blue') #绘制折线图像1,圆形点，标签，线宽
# a2=ax2.scatter(vardata_times, vardata,s=1, c='red') #同上
# plt.legend((a1,a2),('rain',unicode(varname,'utf-8')),scatterpoints=10,loc=1,bbox_to_anchor=(1,1), bbox_transform=ax1.transAxes,prop=font)
a1=ax1.plot(rain_times, raindata, 'b-',label='rain',alpha=0.5)#绘制折线图像1,圆形点，标签，线宽
a2=ax2.plot(vardata_times, vardata, 'r.',label=unicode(varname,'utf-8'),markersize=3,alpha=0.5)#同上
labs = [l.get_label()for lin a1+a2]
ax1.legend(a1+a2,labs,loc=1,bbox_to_anchor=(1,1), bbox_transform=ax1.transAxes,prop=font)
fig.autofmt_xdate()
# plt.xlim(xaxis[0],xaxis[1])
if xaxis!=None:plt.xlim(dt.datetime.strptime(xaxis[0],'%Y%m%d%H%M'),dt.datetime.strptime(xaxis[1],'%Y%m%d%H%M'))
ax1.tick_params(axis='x',labelsize=18,length=10,width=5,direction='out')
ax1.tick_params(axis='y',labelsize=30,colors='blue')
ax2.tick_params(axis='y',labelsize=30,colors='red')
ax1.set_yticks([2, 4, 8,20,50], minor=True);
ax1.grid(True,axis='x');ax1.grid(True, which='minor',color='blue',linestyle='--',linewidth=1)
# plt.gca().xaxis.set_major_formatter(mdates.DateFormatter('%H:%M'))
plt.savefig(unicode(varname,'utf-8')+'_plus_rain'+'.png')
# plt.savefig(unicode(varname,'utf-8')+'_plus_rain'+'.eps')
# plt.savefig('water_plus_rain'+'.eps')
plt.close()
def draw_vardata_2D_and_rain_2yaxis(varname,vardata,vardata_times,raindata,rain_times,xaxis=['201907110000','202006100000']):
# fig=plt.figure(figsize=(10,6))
fig, ax1 = plt.subplots(figsize=(15,9))
ax2=ax1.twinx()
levels=vd.vardic[varname]['levels'];
# if levels[-1]==10000:amplification_factor=20000
if levels[-1]==10000:amplification_factor=10000.0/np.amax(raindata)
# if levels[-1]==2000:amplification_factor=20000/5
if levels[-1]==2000:amplification_factor=2000/np.amax(raindata)
# levels=np.array(levels)/1000.0
# print levels
# exit()
xx,yy=np.meshgrid(vardata_times,levels)
font = FontProperties(fname=r"C:\\Windows\\Fonts\\simsun.ttc", size=45)
# plt.legend((a1,a2),('rain',unicode(varname,'utf-8')),prop=font)
vardata = np.ma.masked_where(np.isnan(vardata),vardata)
a2=ax1.pcolormesh(xx,yy, vardata.transpose(),label=unicode(varname,'utf-8'),cmap='coolwarm')#同上
# a2=ax2.contourf(xx,yy, vardata.transpose()) #同上
a1=ax2.plot(rain_times, np.array(raindata)*amplification_factor, c='k',label='rain',linewidth=1,alpha=0.5)#绘制折线图像1,圆形点，标签，线宽
# ax2.set_ylim(0,np.amax(raindata)*amplification_factor)
# a1=ax1.scatter(rain_times, raindata,s=10, c='blue',label='rain') #绘制折线图像1,圆形点，标签，线宽
ax2.get_yaxis().set_visible(False)
ax2.invert_yaxis()
fig.legend(loc=1,bbox_to_anchor=(1,1), bbox_transform=ax1.transAxes,prop=font)
fig.autofmt_xdate()
plt.xlim(dt.datetime.strptime(xaxis[0],'%Y%m%d%H%M'),
dt.datetime.strptime(xaxis[1],'%Y%m%d%H%M'))
ax1.tick_params(axis='x',labelsize=20,length=10,width=5,labelrotation=90)
ax1.tick_params(axis='y',labelsize=30,colors='k')
ax2.tick_params(axis='y',labelsize=20,colors='blue')
ax1.set_yticks([2, 4, 8,20,50], minor=True);
# ax1.grid(True,axis='x')
# plt.gca().xaxis.set_major_formatter(mdates.DateFormatter('%H:%M'))
plt.gca().xaxis.set_major_formatter(mdates.DateFormatter('%y-%m-%d'))
# divider = make_axes_locatable(ax1)
# cax = divider.append_axes('bottom', size='5%', pad=0.1)
plt.colorbar(mappable=a2)
plt.tight_layout()
plt.savefig(unicode(varname,'utf-8')+'_plus_rain'+'.png')
plt.close()
for varnamein vd.vardic:
# for varname in ['水汽总含量']:
# for varname in ['AI','KI','TTI','LI','SI','LCL','LFC','EC','BLH ']:
# if vd.vardic[varname]['filemark']=='L36000AMT':continue
# if vd.vardic[varname]['levels']==None:continue
# if '廓线' not in varname:continue #### 画廓线
if varnamein ['K通道接收机电压1B', 'V通道接收机电压1B', 'K通道模块噪声源电压1B', 'V通道模块噪声源电压1B', '方位1B', '俯仰1B','V通道亮温1D',
'K通道接收机电压1D', 'V通道接收机电压1D', 'K通道模块噪声源电压1D', 'V通道模块噪声源电压1D', 'K通道模块噪声源温度', 'V通道模块噪声源温度',
'K通道腔体温度', 'V通道腔体温度', 'K通道模块温度', 'V通道模块温度', '方位1D', '俯仰1D', '方位2B', '俯仰2B', '方位2D', '俯仰2D',
'方位3D', '俯仰3D']:continue
# varname
if vd.vardic[varname]['levels']==None:
#### 所有值均为90.0; 多个分号
print varname
# continue
dates_all,data_all=get_data_by_varname_of_micro_and_daterange(varname=varname,startday='20190701',endday='20200610')
times=np.array([dt.datetime(i[0],i[1],i[2],i[3],i[4],i[5])for iin dates_all])#### xaxis
# draw_vardata_2D_and_rain_2yaxis(varname=varname,vardata=data_all,vardata_times=times,raindata=rain_list,rain_times=time_list)
draw_vardata_1D_and_rain_2yaxis(varname=varname,vardata=data_all,vardata_times=times,raindata=data_rain,rain_times=dates_rain)
# if vd.vardic[varname]['levels'] != None and '廓线' in varname:
# ##### 所有值均为90.0; 多个分号
# dates_all, data_all = get_data_by_varname_of_micro_and_daterange(varname=varname, startday='20190701',endday='20200610')
# # dates_all, data_all = get_data_by_varname_of_micro_and_daterange(varname=varname, startday='20190701',endday='20191001')
#
# times = np.array([dt.datetime(i[0], i[1], i[2], i[3], i[4], i[5]) for i in dates_all]) #### xaxis
# # draw_vardata_2D_and_rain_2yaxis(varname=varname,vardata=data_all,vardata_times=times,raindata=rain_list,rain_times=time_list)
# draw_vardata_2D_and_rain_2yaxis(varname=varname, vardata=data_all, vardata_times=times, raindata=data_rain,
# rain_times=dates_rain)
# exit()
# if vd.vardic[varname]['levels']==None:
# draw_1D(varname,times,data_all)
# if vd.vardic[varname]['levels']!=None:
# draw_2D(varname,times,data_all)