Make a pandas dataframe from a dictionary of lists

Question

My question is a bit linked to this one, except that I want to turn my lists into rows not columns, and specify the names of the columns.

For example, I start with a dictionary:

my_dict = {'label1' : [0.1, 1.1, 32.1] , 'label2' : [12.3, 5.3, 2.4], 'label3' : [1.5, 7.5, 7.4]}

I want to be able to specify column names like:

cols = ['labels','val1' ,'val2, 'val3']

and create the dataframe

labels  | val1  | val2  | val3
label1    0.1     1.1     32.1
label2    12.3    5.3     2.4
label3    1.5     7.5     7.4

Answer 1

You want a transposed dataframe and then assign column names,

df = pd.DataFrame(my_dict)
df = df.T.reset_index()
df.columns = cols

    labels  val1    val2    val3
0   label1  0.1     1.1     32.1
1   label2  12.3    5.3     2.4
2   label3  1.5     7.5     7.4

Answer 2

You can use pandas.DataFrame.from_dict but using 'orientation' of index rather than columns.

This allows the labels to be used as the index.

df = pd.DataFrame.from_dict(my_dict, orient='index', columns =  ['val1' ,'val2', 'val3'])
df.index.name = 'labels'  # rename index column

print(df)

Output

      val1  val2  val3
labels                  
label1   0.1   1.1  32.1
label2  12.3   5.3   2.4
label3   1.5   7.5   7.4