The integer index or name(s) of the column level to convert into a row level. By default, level=-1, which means that the inner-most column level is converted.

2. dropnalink | boolean | optional

Whether or not to drop resulting rows that contain just NaN. By default, dropna=True.

Return Value

A Series or a DataFrame.

Examples

Stacking single-level DataFrames

Consider the following single-level DataFrame:


        
        
            
                
                
                    df = pd.DataFrame([[2,3],[4,5]], columns=["alice","bob"], index=["age","height"])
df
                
            
                   alice   bob
age      2      3
height   4      5

Calling stack() on df gives:


        
        
            
                
                
                    df.stack()
                
            
            age     alice    2
        bob      3
height  alice    4
        bob      5
dtype: int64

Here, note the following:

the return type is Series with a 2-level index.
the row labels and the column labels in df have merged to form a multi-index.

Stacking DataFrames with multi-level columns

Consider the following DataFrame with multi-level columns:


        
        
            
                
                
                    index = [("A", "alice"), ("A", "bob"), ("B","cathy")]
multi_index = pd.MultiIndex.from_tuples(index)
df = pd.DataFrame([[2,3,4],[5,6,7]], columns=multi_index, index=["age","height"])
df
                
            
                    A             B
        alice   bob   cathy
age     2       3     4
height  5       6     7

By default, level=-1, which means that the inner-most column level ([alice,bob,cathy]) will be converted into a row level:


        
        
            
                
                
                    df.stack()
                
            
                            A     B
age     alice   2.0   NaN
        bob     3.0   NaN
        cathy   NaN   4.0
height  alice   5.0   NaN
        bob     6.0   NaN
        cathy   NaN   7.0

Note the following:

the inner-most column level ([alice, bob, cathy]) became a row index, and is positioned as the inner-most level.
stacking columns with multi-levels often yield many NaN since, for instance, no data exists about the age of alice in group B.

To specify which levels to convert, pass the level parameter like so:


        
        
            
                
                
                    df.stack(level=0)
                
            
                         alice  bob   cathy
age      A   2.0    3.0   NaN
         B   NaN    NaN   4.0
height   A   5.0    6.0   NaN
         B   NaN    NaN   7.0

Here, level=0 means that that outermost column level ([A,B]) is converted into a row level.

Specifying dropna

Consider the following DataFrame:


        
        
            
                
                
                    index = [("A", "alice"), ("A", "bob"), ("B","cathy")]
multi_index = pd.MultiIndex.from_tuples(index)
df = pd.DataFrame([[2,3,None],[5,6,7]], columns=multi_index, index=["age","height"])
df
                
            
                    A            B
        alice  bob   cathy
age     2      3     NaN
height  5      6     7.0

By default, dropna=True, which means that rows that contain just NaN will be removed from the result:


        
        
            
                
                
                    df.stack()
                
            
                           A     B
age     alice  2.0   NaN
        bob    3.0   NaN
height  alice  5.0   NaN
        bob    6.0   NaN
        cathy  NaN   7.0

Notice how cathy's row for the age level is missing. This is because it only contains NaN.

To keep all rows, pass dropna=False like so:


        
        
            
                
                
                    df.stack(dropna=False)
                
            
                    A             B
age     alice   2.0   NaN
        bob     3.0   NaN
        cathy   NaN   NaN
height  alice   5.0   NaN
        bob     6.0   NaN
        cathy   NaN   7.0