CODEWITHSUNDEEP
pythonpandas
user11877626
user11877626

Reputation:

how to map 1 and 0 for all true and false

I have two column in dataframe Column_1 and Column_2 if both contain same value then marked it as 1 as 0

I tried this but it return list of size 1 and inside list it contain series of size 416

    splitsE = [(df.COLUMN_1.str.contains(' Each')) & 
        (df.COLUMN_2.str.contains(' EACH'))]

i even tried this one also but it doesnt map

      splitsE = [(df.COLUMN_1.str.contains(' Each')) & 
        (df.COLUMN_2.str.contains(' EACH'))]
    splitsE = list(map(list, zip(*splitsE)))
   df['CASE#'] = pd.Series(splitsE).map(({True:'1', False:'0'}))

if i want to update df['CASE#'] where it contain / as 2 in place of 0

     COLUMN_3  CASE#  
    25/PACK      0  
    EACH         1 
    100/BOTTLE   0  
    25/PACK      0 
     NaN         0
    3/PACK       0 
    EACH         1

Upvotes: 1

Views: 729

Answers (2)

Walid Bousseta
Walid Bousseta

Reputation: 1469

If I correctly understand your question, here is the solution:

import pandas as pd
data =  [[1,2], [2,5], [3,3], [4,7],[9,9], [6,5]]
df = pd.DataFrame(columns=['col1', 'col2'], data=data)
df.head()

df.head() table

df['comp'] = 1 * (df['col1'] == df['col2'])
df.head()

df.head() table 2

Upvotes: 0

jezrael
jezrael

Reputation: 862511

Use:

splitsE = (df.COLUMN_1.str.contains(' Each')) & (df.COLUMN_2.str.contains(' EACH'))

I think simpliest is convert boolean mask to integer for True/False to 1/0 map:

df['CASE#'] = splitsE.astype(int)

Another solution is use numpy.where:

df['CASE#'] = np.where(splitsE, 1, 0)

Or map with dictionary with removed one ():

df['CASE#'] = splitsE.map({True:'1', False:'0'})

EDIT: For another condition use numpy.select:

mask1 = df.COLUMN_3.str.contains('EACH', case=False, na=False)
mask2 = df.COLUMN_3.str.contains('/', case=False, na=False)

df['CASE#'] = np.select([mask1, mask2], [1, 2], default=0)
print (df)
     COLUMN_3  CASE#
0     25/PACK      2
1        EACH      1
2  100/BOTTLE      2
3     25/PACK      2
4         NaN      0
5      3/PACK      2
6        EACH      1

Upvotes: 2

Related Questions