How to select specific rows based on specific condition-CodePudding

I am working in Pandas Python. I am trying to select specific rows based on specific condition. From the below dataset, I want the system groups which has Type 1 in it. System groups which don't have type 1 can be ignored.

System	Type
A	1
A	2
A	2
A	3
B	1
B	2
C	2
D	3

Required Output

System	Type
A	1
A	2
A	2
A	3
B	1
B	2

System A and B is obtained in the required output becuase it contain the Type 1 value. C and D groups has been ignore due to no Type 1 in them. I am trying to do with groupby but unable to extend this function to check for presence type 1 in it in the condition. Please help

Code to generate dataframe

import pandas as pd
  
data = [['A', 1], ['A', 2], ['A', 2],['A',3],['B',1],['B',2],['C',2],['C',3],['D',3]]
  
df = pd.DataFrame(data, columns=['System', 'Type'])

CodePudding user response：

df[df['System'].isin(df[df['Type'] == 1]['System'])]

  System  Type
0      A     1
1      A     2
2      A     2
3      A     3
4      B     1
5      B     2

First, you filter the df to only rows with Type=1, and select the System column. Then you filter the df to only includes rows where System is in that list

It might be faster to get the values into a set and search in it

df[df['System'].isin(set(df[df['Type'] == 1]['System'].values))]

CodePudding user response：

let's say the table variable is df . Then the code will be

sys = list(df[df['Type'] == 1]['System'].values)
ans = df[df['System'].isin(sys)]

ans is your preferred table. I am sure there are better ways but hopefully this works.

CodePudding user response：

You can use `where clause` it will be much faster than conditional selection

selectedvariables=df.where(df['Type']==1)["System"].dropna()
print(df[df["System"].isin(selectedvariables)])

You can use where clause it will be much faster than conditional selection

You can use `where clause` it will be much faster than conditional selection