pandas python: adding blank columns to df

Question

I'm trying to add x number of blank columns to a dataframe.

Here is my function:

def fill_with_bars(df, number=10):
    '''
    Add blank, empty columns to dataframe, at position 0
    '''

    numofcols = len(df.columns)

    while numofcols < number:
        whitespace = ''
        df.insert(0, whitespace, whitespace, allow_duplicates=True)
        whitespace += whitespace
    return df

but I get this error

ValueError: Wrong number of items passed 2, placement implies 1

I'm not sure what I'm doing wrong?

This seems very wasteful why not just construct a df with the correct number of additional columns and then concat? — EdChum
– EdChum, Commented Jun 17, 2015 at 11:00
I think the problem is that on the third iteration the code barfs because although you allow duplicates, you now have 2 columns with empty strings as column names but this then fails the internal checks as it's expecting a unique column to append but it now finds 2, if your column names were unique then this would work but again, why not just construct a df with the correct dimensions and concat once — EdChum
– EdChum, Commented Jun 17, 2015 at 11:06
@EdChum - this is why I add an extra white space for each column name so it thinks its unique. — Boosted_d16
– Boosted_d16, Commented Jun 17, 2015 at 11:10
I cannot construct a df with the correct dimensions because of the process I working in. — Boosted_d16
– Boosted_d16, Commented Jun 17, 2015 at 11:11
is there another way I can add empty columns to an existing df? — Boosted_d16
– Boosted_d16, Commented Jun 17, 2015 at 11:15

EdChum · Accepted Answer · 2015-06-17 12:22:24Z

Rather than inserting a column at a time, I'd create a df of the dimensions you desired and then call concat:

In [72]:
def fill_with_bars(df, number=10):
    return pd.concat([pd.DataFrame([],index=df.index, columns=range(10)).fillna(''), df], axis=1)

df = pd.DataFrame({'a':np.arange(10), 'b':np.arange(10)})
fill_with_bars(df)

Out[72]:
  0 1 2 3 4 5 6 7 8 9  a  b
0                      0  0
1                      1  1
2                      2  2
3                      3  3
4                      4  4
5                      5  5
6                      6  6
7                      7  7
8                      8  8
9                      9  9

As to why you get that error:

It's because your str is not a single whitespace, it's an empty string:

In [75]:
whitespace = ''
whitespace + whitespace
Out[75]:
''

So on the 3rd iteration, it's trying to lookup the column, expecting there to be only a single column but there are 2 so it failes the internal checks as it now finds 2 columns called ''.

score 1 · Accepted Answer · 2015-06-17 12:08:37Z

1

Try this:

def fill_with_bars(old_df, number=10):
    empty_col = [' '*i for i in range(1,number+1)]
    tmp = df(columns=empty_col)
    return pd.concat([tmp,old_df], axis=1).fillna('')

edited Jun 17, 2015 at 12:08

answered Jun 17, 2015 at 11:25

user4999547

1 Comment

Boosted_d16 Over a year ago

The values come up as NaN, is there a way to use a blank string instead?

Collectives™ on Stack Overflow

pandas python: adding blank columns to df

2 Answers 2

Comments

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related