第七色在线视频,2021少妇久久久久久久久久,亚洲欧洲精品成人久久av18,亚洲国产精品特色大片观看完整版,孙宇晨将参加特朗普的晚宴

'''

Created on 2018年7月10日

Kaggle網(wǎng)站公開的數(shù)據(jù)

@author: Administrator

'''

#-*-coding:utf-8-*-

from __future__ import division

import pandas as pd

import numpy as np

from scipy import stats

##0.Read Data ##

df = pd.read_csv("train.csv")

label = df['TARGET']

df = df.drop(['ID', 'TARGET'], axis = 1)

##1.Basic Analysis##

#(1)Missing Value#

missSet = [np.nan, 9999999999, -999999]

#(2)Count distinct#

len(df.iloc[:,0].unique())

count_un = df.iloc[:,0:3].apply(lambda x:len(x.unique()))

#(3)Zero Values#

np.sum(df.iloc[:,0]==0)

count_zero = df.iloc[:,0:3].apply(lambda x:np.sum(x==0))

?

#(4)Mean Values#

np.mean(df.iloc[:,0]) #沒有去除缺失值之前的均值很低

df.iloc[:,0][~np.isin(df.iloc[:,0],missSet)]#去除缺失值

np.mean(df.iloc[:,0][~np.isin(df.iloc[:,0],missSet)])#去除缺失值后進(jìn)行均值計(jì)算

df_mean = df.iloc[:,0:3].apply(lambda x:np.mean(x[~np.isin(x, missSet)]))

? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ??

#(5)Median Values#

np.median(df.iloc[:,0])#沒有去除缺失值之前

?

df.iloc[:,0][~np.isin(df.iloc[:,0], missSet)]#去除缺失值

np.median(df.iloc[:,0][~np.isin(df.iloc[:,0],missSet)])#去除缺失值后進(jìn)行中位數(shù)計(jì)算

?

df_median = df.iloc[:,0:3].apply(lambda x:np.median(x[~np.isin(x, missSet)])) ? ? ? ? ? ? ? ? ? ? ? ? ?

?

#(6)Mode Values#

df_mode = df.iloc[:,0:3].apply(lambda x:stats.mode(x[~np.isin(x, missSet)])[0][0])

? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ??

#(7)Mode Percentage#

df_mode_count = df.iloc[:,0:3].apply(lambda x:stats.mode(x[~np.isin(x, missSet)])[1][0])

df_mode_perct = df_mode_count/df.shape[0]

#(8)Min Values#

np.min(df.iloc[:,0])

df.iloc[:,0][~np.isin(df.iloc[:,0], missSet)]#去除缺失值

np.min(df.iloc[:,0][~np.isin(df.iloc[:,0],missSet)])#去除缺失值后進(jìn)行最小值計(jì)算

?

df_min = df.iloc[:,0:3].apply(lambda x:np.min(x[~np.isin(x, missSet)])) ?

#(9)Max Values#

np.max(df.iloc[:,0])

df.iloc[:,0][~np.isin(df.iloc[:,0], missSet)]#去除缺失值

np.max(df.iloc[:,0][~np.isin(df.iloc[:,0],missSet)])#去除缺失值后進(jìn)行最大值計(jì)算

?

df_max = df.iloc[:,0:3].apply(lambda x:np.max(x[~np.isin(x, missSet)]))

#(10)quantile Values#

np.percentile(df.iloc[:,0],(1,5,25,50,75,95,99))

df.iloc[:,0][~np.isin(df.iloc[:,0], missSet)]#去除缺失值

np.percentile(df.iloc[:,0][~np.isin(df.iloc[:,0], missSet)], (1,5,25,50,75,95,99))#去除缺失值后進(jìn)行分位點(diǎn)計(jì)算

? ?

json_quantile ={}

for i,name in enumerate(df.iloc[:,0:3].columns):

? ? print('the {} columns:{}').format(i,name)

? ? json_quantile[name] = np.percentile(df[name][~np.isin(df[name],missSet)],(1,5,25,50,75,95,99))

df_quantile = pd.DataFrame(json_quantile)[df.iloc[:,0:3].columns].T

#(11)Frequent Values#

df.iloc[:,0].value_counts().iloc[0:5,] ??

df.iloc[:,0][~np.isin(df.iloc[:,0],missSet)]#去除缺失值

df.iloc[:,0][~np.isin(df.iloc[:,0], missSet)].value_counts()[0:5]#去除缺失值后進(jìn)行頻數(shù)的計(jì)算

json_fre_name =0

json_fre_count ={}

def fill_fre_top_5(x):

? ? if(len(x))<=5:

? ? ? ? new_array =np.full(5, np.nan)

? ? ? ? new_array[0:len(x)] =x

? ? ? ? return new_array

df['ind_var1_0'].value_counts()

df['imp_sal_var16_ultl'].value_counts()

for i,name in enumerate(df[['ind_var1_0','imp_sal_var16_ultl']].columns):

? ? ##1.Index Name

? ? index_name = df[name][~np.isin(df[name],missSet)].value_counts().iloc[0:5, ].index.values

? ? ##1.1 If the length of array is less than 5

? ? index_name=fill_fre_top_5(index_name)

? ??

? ? ##2.Value Count

? ? values_count=df[name][~np.isin(df[name],missSet)].value_counts().iloc[0:5, ].values

? ? ##2.1 If the length of array is less than 5

? ? values_count = fill_fre_top_5(values_count)

? ??

? ? json_fre_count[name]=values_count

? ??

df_fre_name =pd.DataFrame(json_fre_name)[df[['ind_var1_0','imp_sal_var16_ultl']].columns].T

df_fre_count =pd.DataFrame(json_fre_count)[df[['ind_var1_0','imp_sal_var16_ultl']].columns].T

df_fre = pd.concat([df_fre_name, df_fre_count],axis=1)

#(12)Miss Values#

np.sum(np.isin(df.iloc[:,0],missSet))#統(tǒng)計(jì)缺失值

df_miss = df.iloc[:,0:3].apply(lambda x:np.sum(np.isin(x,missSet)))#遍歷每一個遍歷的缺失值情況

基本描述統(tǒng)計(jì)值

特殊值描述值

缺失值

這個沒怎么看懂

數(shù)據(jù)的讀取

導(dǎo)入division是為了防止分母為0

這個

#分位點(diǎn)函數(shù)兩個參數(shù)：數(shù)據(jù)和分位值