2

I have a dataframe filled with several columns. I need to change the values of a column for data normalization like in the following example:

User_id   
751730951     
751730951
0
163526844
...and so on

I need to replace every value in the column that is not 0 (string) in a into something like "is not empty". I have tried it now for hours but still cannot change every value that is not 0 into something else. Replace()-function don't work really good for that. Some good ideas?

EDIT (my solution):

finalResult.loc[finalResult['update_user'] == '0', 'update_user'] = 'empty'
finalResult.loc[finalResult['update_user'] != 'empty', 'update_user'] = 'not empty'
Improve this question
1
  • So df['User_id'] = df['User_id'].replace('0', 'is not empty') does not work? Commented Oct 19, 2017 at 12:47

3 Answers 3

Reset to default
5 
df.loc[df['mycolumn'] != '0', 'mycolumn'] = 'not empty'

or if the value is an int, 

df.loc[df['mycolumn'] != 0, 'mycolumn'] = 'not empty'

df.loc[rows, cols] allows you to get or set a range of values in your DataFrame. First parameter is rows, in which case I'm using a boolean mask to get all rows that don't have a 0 in mycolumn. The second parameter is the column you want to get/set. Since I'm replacing the same column I queried from, it is also mycolumn.

I then simply using the assignment operator to assign the value of 'not empty' like you wanted.

New column containing 'not empty'

If you want a new column to contain the 'not empty' so you're not contaminating your original data in mycolumn, you can do:

df.loc[df['mycolumn'] != 0, 'myNewColumnsName'] = 'not empty'
Improve this answer
Sign up to request clarification or add additional context in comments.

Comments

4

Simpliest is use:

df['User_id'] = df['User_id'].replace('0', 'is not empty')

If 0 is int:

df['User_id'] = df['User_id'].replace(0, 'is not empty')
Improve this answer

Comments

1

Suppose we use a Series with the data specified in the question, named user_id, with a single line you do what you need:

user_id.where(user_id == 0).fillna('is not empty')

I don't like loc very much since I think it complicates the reading.

It might be better than replace because it allows the opposite case:

user_id.where(user_id != 0).fillna('is empty')
Improve this answer

1 Comment

Code only answers are discouraged. Can you describe how this is different and any key reasons it would produce the desired result?

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.