Python. Delete "Unnamed: 0" header lable in dataframe and shift the rest of the header lables to the left (without altering values)

Question

I have a cross-tabulation .txt file of this shape:

	A	B	C	D
A			9.7
B				9.7
C	9.7			9.7
D		9.7	9.7

However, when I try to read it using

df = pd.read_csv(file, sep='\t', index_col=0)

-where file is the location of this .txt file that I'm getting this data frame from-:

	Unnamed: 0	A	B	C	D
A			9.7		NaN
B				9.7	NaN
C	9.7			9.7	NaN
D		9.7	9.7		NaN

As you can see, it shifts the column labels to the right by adding a new label ("Unnamed: 0").

At this point, I'd only need to do three things:

1. To delete Unnamed: 0 label in the header.
2. To shift the rest of the labels of the header to the left (to let the data frame recover its original shape).
3. To delete the last column (the one with all NaN values).

As I'm an absolute beginner in Python, I can't find the way to do so, although it's probably a very simple task.

For more information regarding how my .txt file is built -probably the origin of my problem-, here is the python code I'm using to create the cross-tab shown above (the first table pasted):

file_path is the path where I the cross-tab is created,

unique_nodes is a list where I store the unique nodes I work with. They are the column and row names of the cross-tab: A, B, C, D.

node1 and node2 are nodes from that list.

node_weights are the values of the cells in the table (in this case, 9.7)

    with open(file_path, 'w') as file:
        # Write header row
        file.write('\t' + '\t'.join(unique_nodes) + '\n')
        
        # Write rows of cross-tab data
        for node1 in unique_nodes:  
            file.write(node1 + '\t')
            for node2 in unique_nodes:
                file.write(str(node_weights[node1][node2]) + '\t')
            file.write('\n')

Answer 1

You can just shift the column names to the left by 1 and drop the last column.

df.columns = df.columns.to_list()[1:] + ['temp']
df = df.drop('temp', axis = 1)