Python Forum
How replace Unnamed header column with previous - Printable Version

+- Python Forum (https://python-forum.io)
+-- Forum: Python Coding (https://python-forum.io/forum-7.html)
+--- Forum: Data Science (https://python-forum.io/forum-44.html)
+--- Thread: How replace Unnamed header column with previous (/thread-20501.html)

How replace Unnamed header column with previous - SriMekala - Aug-14-2019

Hi,
I have excel file with merged columns as below:

Category	Group			Sub_valid	
	        Name	Item	Value	 Color	 Status
Group1	    VGT	    Power	0	     VGT	 Power
	        BHA	    Voltage	1	     BHA	 Voltage
Group is merged for Name & Item column, Sub_valid is a merged column for Value, Color, Status.
When I read using:
import pandas as pd
df = pd.read_excel('D:PythonCodes\mergedinput.xlsx',sheetname='Sheet1')
The column above Item is "Unnamed0" & Color column & Status column are "Unnamed1", Unnamed2",

But I want:
Category	Group  Group    Sub_valid  Sub_valid  Sub_valid	
	        Name	Item	Value	   Color	  Status
Group1	    VGT	    Power	0	       VGT	      Power
	        BHA	    Voltage	1	       BHA	      Voltage
I use below code but it needs to do multiple time:

df.rename(columns={'*Unnamed: 1':'nan'}, inplace=True)
but I want to replace if header starts with: Unnamed"

RE: How replace Unnamed header column with previous - boring_accountant - Aug-14-2019

Assuming you want to replace all columns containing the string 'Unnamed' with the string 'nan', this will do:
print(df)
df.columns = [col if not 'Unnamed' in col else 'nan' for col in df.columns]
print(df)
Output:
  First  Unnamed: 1  Unnamed: 2
0     A           1           3
1     B           2           4

  First  nan  nan
0     A    1    3
1     B    2    4
As an alternative, the following yields equal results. I'm not sure whichever I prefer but both are possible.
df.columns = df.columns.map(lambda col: col if not 'Unnamed' in col else 'nan')


RE: How replace Unnamed header column with previous - SriMekala - Aug-15-2019 

Category    Group  Group    Sub_valid  Sub_valid  Sub_valid 
            Name    Item    Value      Color      Status
Group1      VGT     Power   0          VGT        Power
            BHA     Voltage 1          BHA        Voltage
But I want to "Unnamed column should replace with previous not "Unnamed column" like, above

RE: How replace Unnamed header column with previous - boring_accountant - Aug-15-2019

The following function should do the trick. This won't win any beauty prize however as I just quickly hacked this together.
def repeat_cols(columns):
    last_non_unnamed = '' if 'unnamed' in columns[0].lower() else columns[0]
    new_columns = []
    new_column = ''
    for column in columns:
        if 'unnamed' in column.lower():
            new_column = last_non_unnamed
        else:
            new_column = column
            last_non_unnamed = new_column
        new_columns.append(new_column)
    return new_columns

df.columns = repeat_cols(df.columns)