Selecting all numerical values in data-frame and converting it to int in panda

Question

I have data like:

I want to make all numerical values to int, no decimal will be here. like this:

I was using code like list:

             df=pd.read_csv("file.csv")

             df = df.astype('int64')

But it was not working. Because it was saying:

          ValueError: Cannot convert NA to integer

Because, I have some nan value inside. There are string also in the column one row 2&3. I think the solution could be, selecting all numerical values from the data frame and converting to int. Can you suggest anything?

the redirected duplicate question seems to be different from this. Here, the problem is not `nan` fill value, but selecting the numerical columns. — Rahul Goswami, Jun 22 '18 at 22:39

score -1 · Answer 1 · answered Jun 22 '18 at 22:37

-1

There are some null values in your data frame. First, you need to covert them to an integer (for e.g. 0 here). use df.fillna for that. Then select all numerical columns.

cols = ["num_col1", "num_col2"]
df[cols] = df[cols].astype("int64")

answered Jun 22 '18 at 22:37

Rahul Goswami

535
3
14

Where does it say 0 is NA? – pyeR_biz Jun 22 '18 at 22:48
Not a great answer to an especially poor question. – cs95 Jun 22 '18 at 22:53
Wait, am I the only one to understand OPs question? It has also been marked as wrong duplicate. Its more close (not exactly) to [this question](https://stackoverflow.com/questions/25039626/how-do-i-find-numeric-columns-in-pandas). – Rahul Goswami Jun 23 '18 at 04:30

Selecting all numerical values in data-frame and converting it to int in panda

1 Answers1