CODEWITHSUNDEEP
pythoncsvpandas
tapzx2
tapzx2

Reputation: 1071

Iterating w/ Pandas DataFrame and Changing Values

I have a pandas dataframe where I want to assign a random number to each row based on a value in the row and write out a dataframe.

So I'm trying:

for index, row in person[person['AGE_R'] == 1].iterrows():
    row = index, random.randint(1, 15)

But I can't quite figure out how to write out a dataframe from it (not possible?). I been able to get out a list of tuples which I might be able to munge into a workable format, but I'm sure there's a better way.

I previously tried:

person[person['AGE_R'] == 1] = random.randint(1, 15)

But that sets all the 1's of 'AGE_R to what ever the randint is. Useful, but not what I'm looking for.

Any suggestions?

Thank you!

Upvotes: 1

Views: 140

Answers (1)

roman
roman

Reputation: 117400

If you want to do vectorized operation, you can use numpy.random.randint:

>>> df = pd.DataFrame({'AGE_R':[1,2,3,5,4,3,1]})
>>> df
   AGE_R
0      1
1      2
2      3
3      5
4      4
5      3
6      1
>>> df.ix[df['AGE_R'] == 1, 'AGE_R'] = np.random.randint(1, 15, len(df[df['AGE_R'] == 1]))
>>> df
   AGE_R
0      5
1      2
2      3
3      5
4      4
5      3
6      11

Or you can use apply:

>>> df.ix[df['AGE_R'] == 1, 'AGE_R'] = df.ix[df['AGE_R'] == 1].apply(lambda x: np.random.randint(1, 15), axis = 1)
>>> df
   AGE_R
0      5
1      2
2      3
3      5
4      4
5      3
6     12

Upvotes: 1

Related Questions