How one can use variable in a question in pandas

Let’s create a pattern dataframe having 3 columns and 4 rows. This dataframe is used for demonstration function.

import pandas as pd
df = pd.DataFrame({"col1" : vary(1,5), 
                   "col2" : ['A A','B B','A A','B B'],
                   "col3" : ['A A','A A','B B','B B']
                   })

Filter a price A A in column col2

With the intention to do reference of a variable in question, that you must use @.

newdf = df.question("col2 == 'A A'")

myval1 = 'A A'
newdf = df.question("col2 == @myval1")

How one can cross column identify as a variable in question

As a substitute of filter worth we’re referring the column which we wish to use for subetting or filtering.

myvar1 = 'col2'
newdf2 = df.question("{0} == 'A A'".format(myvar1))

{0} takes a price of variable myvar1.
"{0} == 'A A'".format(myvar1) returns "col2 == 'A A'"

Incase you wish to cross a number of columns as variables in question. Right here we’re utilizing columns col2 and col3.

myvar1 = 'col2'
myvar2 = 'col3'
newdf2 = df.question("{0} == 'A A' & {1} == 'B B'".format(myvar1, myvar2)) 

"{0} == 'A A' & {1} == 'B B'".format(myvar1, myvar2) is equal to "col2 == 'A A' & col3 == 'B B'"

How one can deal with house in column identify

Let’s rename column col2 by together with an area in between for illustration function.

df.rename(columns={'col2':'col 2'}, inplace = True)

By utilizing backticks `` you may cross a column which accommodates house.

myvar1 = '`col 2`'
newdf = df.question("{0} == 'A A'".format(myvar1))