pandas最快入门（二）-查询与新增

import pandas as pd

df=pd.read_excel('000001.SZ.xlsx',encoding='gbk')

df.head()

#设置索引为编号，方便筛选
df.set_index('trade_date',inplace=True)

#替换掉ts_code后面的后缀
df.loc[:,'ts_code']=df['ts_code'].str.replace('.SZ','').astype('int32')

df.head()

#得到单个值
df.loc[20160108,'open']

11.21

#得到一个series
df.loc[[20160107,20160108],'open']

trade_date
20160107    11.41
20160108    11.21
Name: open, dtype: float64

#得到一个Dataframe
df.loc[[20160107,20160108],['open','low']]

df.loc[20160106:20160108,'open']

trade_date
20160106    11.42
20160107    11.41
20160108    11.21
Name: open, dtype: float64

df.loc[20160107,'open':'low']

open    11.41
high    11.41
low     10.91
Name: 20160107, dtype: float64

df.loc[20160106:20160108,'open':'low']

#bool条件语句
df['open']<11.5

trade_date
20160104    False
20160105     True
20160106     True
20160107     True
20160108     True
            ...  
20201225    False
20201228    False
20201229    False
20201230    False
20201231    False
Name: open, Length: 1218, dtype: bool

#利用条件语句进行查询
#&表示与，每个条件表达式需要加括号
df.loc[(df['open']<11.3) & (df['high']>11.5),:]

#lambda表达式-传入df，返回冒号后面的内容
df.loc[lambda df: (df['open']<11.3) & (df['high']>11.5),:]

#自定义查询函数，返回条件语句
def query_my_data(df):
    return (df.index>20160105) & (df['close']==11.54)

df.loc[query_my_data,:]

#计算当日股票价格变动
df.loc[:,"today_change"]=df['close']-df['open']

df.head()

对series进行操作，需指定axis

def get_change(x):
    if x['today_change']>0:
        return "up"
    else:
        return "down"

#需要设置axis=1，表示series的index是columns
df.loc[:,"当日变动"]=df.apply(get_change,axis=1)

#统计
df['当日变动'].value_counts()

down    621
up      597
Name: 当日变动, dtype: int64

df.head()

df['pct_observe']=''
df.loc[df['pct_chg']>0,'pct_observe']='+'
df.loc[df['pct_chg']<=0,'pct_observe']='-'

df.head()

源代码和相应数据下载请点击 https://github.com/MiaoChenglin125/Data-Acquisition/tree/master/pandas%E6%93%8D%E4%BD%9C