CODEWITHSUNDEEP
python-3.xpandasdataframe
messy748
messy748

Reputation: 327

How can I replace all occurrences of 0 with 1 in my dataframe?

I am working with a dataframe that looks something like

duration count 0 23 2 12 0 43 2 12 0 43 4 65 1 22

I simply want to replace all of the 0's in the duration column with 1. I tried to do this by

df_sorted = df_sorted.replace(df_sorted['duration'] == 0, df_sorted['duration'] == 1)

However, I run this and the dataframe seems unaffected. I do not get an error message and nothing seems to change. I am working with a much bigger dataframe with a lot more rows/columns so maybe I'm missing that something else is changing? Regardless, the 0's still remain after I run the line of code.

Upvotes: 3

Views: 66

Answers (3)

S.N
S.N

Reputation: 5140

You are missing inplace=True which essentially make the change permanent.

Note: inplace=True returns None inplace=False returns a copy of the object with the operation performed.

df_sorted = pd.DataFrame({'duration':[1,0,2,0],'Count':[23,23,2,1]})
df_sorted.replace({'duration': {0:1}}, inplace=True)
print(df_sorted)

Upvotes: 0

piRSquared
piRSquared

Reputation: 294258

short and sweet

df_sorted['duration'] += df_sorted['duration'].eq(0)

df_sorted

   duration  count
0         1     23
1         2     12
2         1     43
3         2     12
4         1     43
5         4     65
6         1     22

Upvotes: 3

Celius Stingher
Celius Stingher

Reputation: 18367

You can use np.where which is somewhat similar to a case when from sql.

df['duration'] = np.where(df['duration'] == 0,1,df['duration'])

Which reads as: Check the value of each row of df['duration'], if it's equal to 0, replace it with a 1, otherwise, return df['duration']

Upvotes: 1

Related Questions