Create pandas DataFrame with more than one column using "data" parameter with lists as inp-CodePudding

I have a problem I couldn't figure out for weeks now but it sounds so simple that i cannot imagine that it's not possible. Assuming that I have data which should be represented in columns, minimum 2 or more. I know there are many ways to create the df and the easiest would be create a dict like

df = pd.DataFrame(
    {
        'col1' : ['a', 'b', 'c'],
        'col2' : ['d', 'e', 'f'],
        'col3' : [1, 2, 3],
    })

but I would like to create it with the syntax:

df = pd.DataFrame(data="here the lists which represent columns", index='...', columns=['...'])

If I have one single list of values as data with index=list('ABCD') and columns=['col1'] it works. The list of data will be a column in the DataFrame with the shape df.shape=(4,1)

If data parameter looks like this:

data = [['a', 'b', 'c'], ['d', 'e', 'f']]

the output will be a df with shape (3,2) because every list will be interpreted as row where 1st row is "a" and "d" and so on, or if index=list('ABC') and columns=[['col1','col2']] will be added then I get the ValueError that "2 columns passed, passed data had 3 columns"

A little workaround would be:

df = pd.DataFrame(data=[['a', 'b', 'c'], ['d', 'e', 'f']], index=['col1', 'col2'])
df = df.T

Is there a way I didn't think of? Change the input of "data" from list to Series or np.array also didn't help for me.

CodePudding user response：

You may want to pass the dict

pd.DataFrame(dict(zip(['col1','col2'],data)))
  col1 col2
0    a    d
1    b    e
2    c    f

CodePudding user response：

You can assign your data to the column header like this:

data = [['a', 'b', 'c'], ['d', 'e', 'f'], [1,2,3]]
df = pd.DataFrame(data={
  'col1': data[0],
  'col2': data[1],
  'col3': data[2]
})