Create a table using dictionary on pandas with specified columns-CodePudding

everyone.

I need a help in creating the table with specified columns

Easy example:

from collections import defaultdict
import pandas as pd
d = defaultdict(list)
mapp = {'A': 'Error 3231', 'B': 'Error 23143243', 'C': 'Error 3242e32'}

for i in ['A', 'B']:
    d[i].append(f'U1_{mapp[i]}')

for i in ['A', 'C']:
    d[i].append(f'R1_{mapp[i]}')

When I use df1 = pd.DataFrame.from_dict(d, orient='index') it creates

                   0              1
A      U1_Error 3231  R1_Error 3231
B  U1_Error 23143243           None
C   R1_Error 3242e32           None

But actually, I want to get another table like the below one:

                   U1                  R1
A       U1_Error 3231       R1_Error 3231
B   U1_Error 23143243                None
C                None    R1_Error 3242e32

The values U1 and R1 are fixed.

Thank you.

CodePudding user response：

Convert d to create your dataframe. It is often preferable to reformat your data before creating your dataframe:

df = pd.Series({(k, v[:2]): v for k, l in d.items() for v in l}).unstack()
print(df)

# Output
                  R1               U1
A  R1_Error dsad2314  U1_Error 232324
B                NaN  U1_Error 232324
C  R1_Error dsad2314              NaN

Transformation:

# From
>>> dict(d)
{'A': ['U1_Error 232324', 'R1_Error dsad2314'],
 'B': ['U1_Error 232324'],
 'C': ['R1_Error dsad2314']}

# To
>>> {(k, v[:2]): v for k, l in d.items() for v in l}
{('A', 'U1'): 'U1_Error 232324',
 ('A', 'R1'): 'R1_Error dsad2314',
 ('B', 'U1'): 'U1_Error 232324',
 ('C', 'R1'): 'R1_Error dsad2314'}

CodePudding user response：

You can create the desired table by transposing the original dataframe and then renaming the columns. Here is an example:

df1 = pd.DataFrame.from_dict(d, orient='index')
df2 = df1.T
df2.columns = ['U1', 'R1']

This will transpose the original dataframe, which will make the columns into rows and vice versa. Then, you can rename the columns to 'U1' and 'R1'.