Transforming a list of tuples into pandas dataframe with the second item as column header and first-CodePudding

I have a following list of tuples: data = [[('abc', 'type1'), ('def', 'type2'), ('ghi', 'type3')], [], [('jkl', 'type4')], [('mno', 'type1'), ('pqr', 'type3')]]

Is it possible to put this into a dataframe using the second item as column header? Desired output:

  type1  type2  type3  type4

0  abc    def    ghi    NaN   

1  NaN    NaN    NaN    NaN

2  NaN    NaN    NaN    jkl

3  mno    NaN    pqr    NaN

CodePudding user response：

This should work:

df = pd.DataFrame(data=None,columns=['type1','type2','type3','type4'])
data = [[('abc', 'type1'), ('def', 'type2'), ('ghi', 'type3')], [], [('jkl', 'type4')], [('mno', 'type1'), ('pqr', 'type3')]]
for n in data:
    datain = [np.NaN, np.NaN, np.NaN, np.NaN]
    for a,b in n:
        if b == 'type1':
            datain[0] = a
        elif b == 'type2':
            datain[1] = a
        elif b == 'type3':
            datain[2] = a
        elif b == 'type4':
            datain[3] = a
    df2 = pd.DataFrame(data=[datain],columns=['type1','type2','type3','type4'])
    df = pd.concat([df,df2])