import pandas as pd
import datetime
import csv
import os
from math import radians,cos,sin,asin,sqrt
#由经纬度计算实际的距离
def geodistance(lng1,lat1,lng2,lat2):   
    lng1, lat1, lng2, lat2 = map(radians, [float(lng1), float(lat1), float(lng2), float(lat2)]) # 经纬度转换成弧度
    dlon=lng2-lng1
    dlat=lat2-lat1
    a=sin(dlat/2)**2 + cos(lat1) * cos(lat2) * sin(dlon/2)**2 
    distance=2*asin(sqrt(a))*6371*1000 # 地球平均半径，6371km
    distance=round(distance/1000,3)
    return distance            #结果的单位为km
#去除两个重复条目
def thesame(data1):
    de=[]
    #data1 = data1['time'].drop_duplicates()
    for i in range(1,len(data1)):
        if data1.iloc[i]["time"]==data1.iloc[i-1]["time"] and data1.iloc[i]["lat"]==data1.iloc[i-1]['lat'] and data1.iloc[i]["lon"]==data1.iloc[i-1]["lon"]:
            de.append(i)
            print(1,data1.iloc[i]["ID"],data1.iloc[i]["time"],data1.iloc[i]["lat"],data1.iloc[i]["lon"])
    for i in range(len(de)):
        data1.drop(de[i],axis=0,inplace=True)
    return data1
#一分钟内坐标不移动的车辆信息删去
def parkcar(data1):
    de=[]
    t=0
    ti=[]
    t1=0
    t2=0
    for i in range(1,len(data1)):
        x1=data.iloc[i]['lat']-data.iloc[i-1]['lat']
        x2=data.iloc[i]['lon']-data.iloc[i-1]['lon']
        if x1==0 and t==0 and x2==0:
            t=1
            t1=data.iloc[i]['time']
        if t==1:
            ti.append(i)
        if t==1 and (x1!=0 or x2!=0):
            t=0
            ti.remove(i)
            t2=data.iloc[i]['time']
            t3=t2-t1
            if t3.seconds>120:
                de.append(ti)
                print(3,data1.iloc[i]["ID"],data1.iloc[i]["time"],data1.iloc[i]["lat"],data1.iloc[i]["lon"])
                ti=[]
    for i in range(len(de)):
        data1.drop(de[i],axis=0,inplace=True)
    return data1
#坐标漂移(去重后才能运行，不然会有时间为零，除数不能为零)
def error1(data1):
    de=[]
    j=0
    d1=[]
    for i in range(1,len(data1)-1):
        if len(data1)==1 or len(data1)==0:
            break
        if i>1 and i<len(data1):
            d1=geodistance(data1.iloc[i-1]['lon'],data1.iloc[i-1]['lat'],data1.iloc[i]['lon'],data1.iloc[i]['lat'])
            d2=geodistance(data1.iloc[i]['lon'],data1.iloc[i]['lat'],data1.iloc[i+1]['lon'],data1.iloc[i+1]['lat'])
            time1=data1.iloc[i]['time']-data1.iloc[i-1]['time']
            time2=data1.iloc[i+1]['time']-data1.iloc[i]['time']
            t1=time1.seconds
            t2=time2.seconds
            if t1==0 or t2==0:
                continue
            else:
                speed2=(d2/t2)*3600
                speed1=(d1/t1)*3600
            if t2<=7200 and speed1>60 and t1<7200 and speed2>60:
                de.append(i)
                print(4,data1.iloc[i]["ID"],data1.iloc[i]["time"],data1.iloc[i]["lat"],data1.iloc[i]["lon"])

    print(de)
    for i in range(len(de)):
        data1.drop(de[i],axis=0,inplace=True)
    return data1
def yuejie(data1):#由于之前对北京市定位的json文件读取不当，没有圈出北京市范围，所以只好大概的定一个范围，不需要再使用上面的insidecity()函数
    de=[]
    for i in range(len(data1)):
        if(data1.iloc[i]["lon"]<115.4 or data1.iloc[i]["lon"]>177.5 or data1.iloc[i]["lat"]<39.433 or data1.iloc[i]["lat"]>41.02):
            de.append(i)
            print(2,data1.iloc[i]["ID"],data1.iloc[i]["time"],data1.iloc[i]["lat"],data1.iloc[i]["lon"])
    for i in range(len(de)):
        data1.drop(de[i],axis=0,inplace=True)
    return data1
def dealwith(data):
    data1=thesame(data)
    data=data.reset_index(drop=True)
    data1=yuejie(data)
    data=data.reset_index(drop=True)
    data1=parkcar(data)
    data=data.reset_index(drop=True)
    return data
path = r"D:\shuju" #文件夹目录
files= os.listdir(path) #得到文件夹下的所有文件名称
for file in files: #遍历文件夹
    position = path+'\\'+ file #构造绝对路径，"\\"，其中一个'\'为转义符           
    with open(position, "r",encoding='utf-8') as f:
        data=pd.read_csv(f,names=['ID','time','lon','lat'])
        data['time']=pd.to_datetime(data['time'])
        data = data.sort_values(by=['time'])    
        data=dealwith(data)
    with open(position,"w",encoding='utf-8') as f:
        for line in data.values:
            f.write((str(line[0])+','+str(line[1])+','+str(line[2])+','+str(line[3])+'\n'))