Files
PriceSystem/PriceSystem/价格涨跌幅大处理/from_calc_to_avg_0604.py
T
2018-06-19 11:20:24 +08:00

211 lines
4.4 KiB
Python

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Mon May 28 17:26:56 2018
@author: luogan
"""
import pymongo
import pytz
import pandas as pd
import numpy
import collections
import pandas
import time
def get_weekend(tt):
from dateutil.parser import parse
bb=parse(tt)
pacific = pytz.timezone('Asia/Shanghai')
return pacific.localize(bb)
weekend = get_weekend('20180309')
client=pymongo.MongoClient('192.168.10.182',27017)
db=client.fangjia.district_stat_0604
def data(ct,rl,lk):
query={'weekend':{'$gte':weekend},'city':ct,'region':rl,'district_name':lk}
#query={'weekend':weekend,'city':'上海'}
ll=list(db.find(query))
pl=pd.DataFrame(ll)
return pl
#a=pd.read_excel('0529.xls')
#a1=a[a['出现问题次数']>1]
def k_means(pp1):
from sklearn.cluster import KMeans
pv = list(pp1)
if len(set(pv)) > 3:
gf = numpy.array([pv]).T
estimator = KMeans(n_clusters=3) # 构造聚类器
estimator.fit(gf) # 聚类
label_pred = estimator.labels_ # 获取聚类标签
aa = collections.Counter(label_pred)
v = pandas.Series(aa)
gg = list(v)
index_max = gg.index(max(gg))
centroids = estimator.cluster_centers_ # 获取聚类中心
# inertia = estimator.inertia_ # 获取聚类准则的总和
center = centroids[index_max][0]
return (float(center))
else:
return float(pp1.mean())
def modify(df,k,km_price):
#print(k)
this_value=df['avg_price'][k]
rat=abs(this_value-km_price)/min(this_value,km_price)
if rat>0.08:
df['avg_price'][k]=km_price
return df
else:
return df
def mm(df,j,km_price):
if j==0:
df1=modify(df,j,km_price)
df2=modify(df1,j+1,km_price)
return df2
elif j==df.shape[0]-1:
df0=modify(df,j-1,km_price)
df1=modify(df0,j,km_price)
#df2=modify(df1,j+1,km_price)
return df1
else:
print('#'*10)
df=modify(df,j-1,km_price)
df=modify(df,j,km_price)
df=modify(df,j+1,km_price)
return df
'''
cl=pymongo.MongoClient('192.168.10.182',27017)
db182=cl.lg.district_stat
'''
'''
for i in a.index:
b1=a.iloc[i]
ct=b1['city']
rl=b1['region']
lk=b1['district_name']
print(ct,rl,lk)
modify_and_save(ct,rl,lk)
'''
def modify_and_save(ct,rl,lk):
print(ct,rl,lk)
df=data(ct,rl,lk)
#df_out=df[(df['markup']>1000) |(df['markup']<-1000) ]
km_price= k_means(df['calc_value'].dropna())
'''
if df_out.shape[0]>0:
for j in df_out.index:
df= mm(df,j,km_price)
#print(df)
#time.sleep(3)
'''
for n in df.index:
df=modify(df,n,km_price)
df=df.sort_values(by='weekend',ascending=False)
for m in range(df.shape[0]-1):
p1=df.iloc[m]
price=p1['avg_price']
_id=p1['_id']
def mark_up(df,m):
p2=df.iloc[m+1]
price_2=p2['avg_price']
t1=p1['weekend']
t2=p2['weekend']
delta=t1-t2
print(delta.days)
if abs(delta.days)==7:
make_up = 10000 * (price - price_2) / price_2
return int(make_up)
else:
return int(0)
mp=mark_up(df,m)
print(mp)
if abs(mp)>1000:
print(mp)
df['avg_price'].iloc[m]=km_price
p2o=df.iloc[m+1]
price_2o=p2o['avg_price']
mpo=10000 * (km_price - price_2o) / price_2o
df['markup'].iloc[m]=mpo
else:
df['markup'].iloc[m]=mp
#print('df=',df)
for q in range(df.shape[0]-1):
avg=df['avg_price'].iloc[q]
mmp=int(df['markup'].iloc[q])
_id=df['_id'].iloc[q]
db.update(
{"_id": _id},
{
"$set": {'avg_price':int(avg),'markup':mmp,
"weekend_markup":mmp,"changed_markup" :mmp }
},True
)