How to create new column from given data in pandas?

Question

I am trying to make a new dataframe of 2 columns. First column is id and second column is predictions. I want to create a dataframe with first column id which is taken from testdata and predictions is an array of 0's and 1's. Column names should be id and predictions.

Sample data:

My new create dataframe should be like this:

Expected output:

id      predictions
3242        0
3246        1
7655        1
1324        0
4643        1

I tried following code but I am getting some error please check screenshot:

@jezrael Check my edited question. I have added testdata and expected output which I need to submit as a csv file. — stone rock, May 28 '18 at 06:57
@jezrael I want to create a new dataframe with 2 columns 1) `id` 2) `predictions`. `id` column should have `id` of testdata and `predictions` see in screenshot, should be the second column. Also both columns should have name `id` and `predictions`. — stone rock, May 28 '18 at 07:02

jezrael · Accepted Answer · 2018-05-28T07:04:10.990

2

I believe need if lenght of testdata is same as predictions:

Test it:

print (len(testdata))

print (len(predictions))

df = pd.DataFrame({'id':testdata['id'], 'predictions':predictions})

edited May 28 '18 at 07:04

answered May 28 '18 at 07:02

jezrael

822,522
95
1,334
1,252

I am getting `ValueError: array length 31349 does not match index length 31350` – stone rock May 28 '18 at 07:04
@stonerock - so need remove last value from `predictions` for same length? – jezrael May 28 '18 at 07:05
@stonerock - if yes, use `df = pd.DataFrame({'id':testdata['id'], 'predictions':predictions[:-1]})` – jezrael May 28 '18 at 07:06
Now I get ValueError: array length 31348 does not match index length 31350 – stone rock May 28 '18 at 07:07
1

@stonerock - so need `df = pd.DataFrame({'id':testdata['id'].values[:-1], 'predictions':predictions})` – jezrael May 28 '18 at 07:08

score 1 · Answer 2 · answered May 28 '18 at 07:05

1

If the length of the arrays are equal

df = pd.DataFrame({'id':testdata['id']})
df["Predictions"] = predictions

answered May 28 '18 at 07:05

Vishnu Kiran

620
7
15

How to create new column from given data in pandas?

2 Answers2