Pandas基础教程 数据读取 索引与计算 数据预处理 自定义函数

Pandas数据处理基于numpy ，DateFram 由多个series组成，每个series代表一行或者一列。
1.Pandas数据读取
import pandas
food_info = pandas.read_csv('food_info.csv')
print(type(food_info)) #数据类型 DateFram
print(food_info.dtypes)
print(help(pandas.read_csv))

food_info.head(3)   #显示数据 显示前3条数据

food_info.tail(4)  #显示后四行

print(food_info.columns)  #显示列名
print(food_info.shape)  #8618个数据，36列指标

print(food_info.loc[0])  #读取第一行数据
food_info.loc[3:6]     #读取3-6行数据

ndb_col = food_info['NDB_No']   #取NDB_No列
print(ndb_col)

columns = ['Zinc_(mg)','Copper_(mg)']   #取'Zinc_(mg)','Copper_(mg)'两列数据
zinc = food_info[columns]
print(zinc)

col_name = food_info.columns.tolist()   #取第一列的表示元素
print(col_name)                    #输出第一列中所有元素

gram_columns = []

for c in col_name:               #找到第一列元素中含有（g）的元素，输出出来
if c.endswith('(g)'):
gram_columns.append(c)
gram_df = food_info[gram_columns]
print(gram_df.head(3))

print(food_info['Iron_(mg)'])    #将Iron_(mg)中的元素都除以1000
div_1000 = food_info['Iron_(mg)']/1000
print(div_1000)

iron_grams = food_info['Iron_(mg)']/1000    #添加Iron_(g) 一列数据
print(food_info.shape)
food_info['Iron_(g)'] = iron_grams
print(food_info.shape)

donkey = food_info['Iron_(mg)'].max()  #找到Iron_（mg）里面的最大值
print(donkey)

food_info.sort_values('Sodium_(mg)',inplace = True)    #升序
print(food_info['Sodium_(mg)'])

food_info.sort_values('Sodium_(mg)',inplace = True,ascending=False)  #降序
print(food_info['Sodium_(mg)'])

2.索引的计算和数据预处理

import pandas as pd   #读取文件
import numpy as np
titanic_survival = pd.read_csv('titanic_train.csv')
titanic_survival.head()

age = titanic_survival['Age']   #输出0-10个age
print(age.loc[0:10])
age_is_null = pd.isnull(age)    #判断age是否为空
print(age_is_null)
age_true_null = age[age_is_null]     #输出age为空的元素
print(age_true_null)
age_true_num = len(age_true_null)   #统计age为空的数量
print(age_true_num)

mean_age = titanic_survival['Age'][age_is_null == False]   #计算年龄的平均值
correct_mean_age = sum(mean_age)/len(mean_age)
print(correct_mean_age)

correct_mean_age = titanic_survival['Age'].mean()  #计算年龄的平均值
print(correct_mean_age)

passager = [1,2,3]   #求1 2 3等座的平均票价
gares = {}
for this_class in passager:
pclass = titanic_survival[titanic_survival['Pclass'] == this_class]
fares = pclass['Fare']
mean = fares.mean()
gares[this_class] = mean
print(gares)

donkey = titanic_survival.pivot_table(index='Pclass',values = 'Fare',aggfunc = np.mean)  #求1 2 3等座的平均票价  简便方法
print(donkey)

donkey = titanic_survival.pivot_table(index='Pclass',values = ['Fare','Age'],aggfunc = np.mean)  #求1 2 3等座的平均票价和年龄
print(donkey)

drop1 = titanic_survival.dropna(axis=1)   #删除列缺失的表  通常不会这样操作因为会丢失变量

drop2 = titanic_survival.dropna(axis=0,subset=['Age','Sex'])  #删除行缺失的元素
print(drop2)

dingwei = titanic_survival.loc[83,'Age']  #定位第83的age是多大
print(dingwei)

#排序
donkey = titanic_survival.sort_values('Age',ascending=False)   #降序排序
print(donkey[0:10])
donkey1 = donkey.reset_index(drop=True)   #从0开始重新设置前面的排序数字
print(donkey1.loc[0:10])

3自定义函数def hundre(colume): #输出第100行所有的数值
donkey = colume.loc[99]
return donkey
hundre = titanic_survival.apply(hundre)
print(hundre)

def donkey(colume):    #输出数值为空的每个变量里面的数量
donkey1 = pd.isnull(colume)
null = colume[donkey1]
return len(null)
count = titanic_survival.apply(donkey)
print(count)

def which_class(row):    #每一行数据中的座位等级
pclass = row['Pclass']
if pd.isnull(pclass):
return 'Unknuwn'
elif pclass == 1:
return 'First'
elif pclass ==2:
return 'Scond'
else:
return 'Third'
classes = titanic_survival.apply(which_class,axis=1)
print(classes)

import pandas as pd
donkey = pd.read_csv('fandango_score_comparison.csv')
film = donkey['FILM']
print(type(film))
print(film.loc[0:5])
donkey1 = donkey['RottenTomatoes']
print(donkey1[0:5])

from pandas import Series
donkey = film.values     #将FILM中的值赋值给donkey
print(type(donkey))     #numpy类型
print(donkey)
donkey2 = donkey1.values   #将RottebTomatoes的值赋给donkey2
print(donkey2)