[FIXED] Split pandas dataframe

Issue

I have a sample dataframe of company A's electronic consumption shown below

year-month	company	GWh
2017-01	A	100
2018-02	A	110
2019-01	A	90
2019-02	A	105
2020-01	A	117
2020-02	A	120

i would like to remove data of year 2020 and split the remaining dataframe into two sets:

1. Train dataset contains records before year 2019
2. Test dataset contains only 2019 records

Solution

Coerce them into datetime and select as required. This is most preferable if you'll need to use them for time analysis

df['year-month']=pd.to_datetime(df['year-month']).dt.strftime('%Y-%m')
df1=df[df['year-month'].lt('2019')]
df2=df[df['year-month'].eq('2019')]

Following your comment, I believe thats a bug. I would edit it to

df['year-month']=pd.to_datetime(df['year-month'])
df1=df[df['year-month'].dt.strftime('%Y').lt('2019')]
df2=df[df['year-month'].dt.strftime('%Y').eq('2019')]

Answered By - wwnde