Whether or not to create a new copy of the source DataFrame. Note that a copy represents a new DataFrame, that is, modifying the transpose would not mutate the original DataFrame and vice versa. By default, copy=False.

WARNING

The source DataFrame is always copied if any of its columns contains mixed types (e.g. a mix of strings and numerics).

Return Value

The transposed DataFrame.

Examples

Basic usage

Consider the following DataFrame:


        
        
            
                
                
                    df = pd.DataFrame({"A":[3,4],"B":[5,6]})
df
                
            
               A  B
0  3  5
1  4  6

Taking its transpose gives:


        
        
            
                
                
                    df.transpose()
                
            
               0  1
A  3  4
B  5  6

Notice how the rows and columns are swapped.

Case when columns contain mixed types

Consider the following DataFrame:


        
        
            
                
                
                    df = pd.DataFrame({"A":[3,4],"B":[True,6]})
df
                
            
               A  B
0  3  True
1  4  6

Here, column B contains mixed-types: a boolean and a numeric.

Computing the transpose gives:


        
        
            
                
                
                    df.transpose()
                
            
               0     1
A  3     4
B  True  6

At first glance, you'd think the new column 1 would be of type numeric. However, this isn't the case:


        
        
            
                
                
                    df.transpose().dtypes
                
            
            0    object
1    object
dtype: object

We see that column 1 is actually of type object, instead of int. This is because the transpose of a DataFrame whose column(s) contain mixed types will have all of its columns as type object.

Pandas DataFrame | T property

A property that represents the transposed DataFrame.

chevron_right