CODEWITHSUNDEEP
pythonpython-3.xpandasdataframemerge
Solar
Solar

Reputation: 455

Pandas giving KeyError when merging on df2.columns

I've got two dataframes. First one is empty but with columns defined:

Empty DataFrame
Columns: [ID, 3120, 3121, 3122, 3123, 3124, 3125, 3126, 3127, 3128, 3129, 3130, 3131, 3146, 3147, 3148, 3149, 3150, 3151, 3152, 3153, 3154, 3155, 3156, 3157]
Index: []

Second dataframe is:

    3123    3124    3125    3126    3127
0   A       B       C       D

Later, I will have another dataframe that will be:

    3146    3147    3148    3149    3150
0   X       Y       Z

And so on. What I want is to put all this little dataframes in the first one to get something like:

ID  3120    3121    3122    3123    3124    3125    3126    3127    3128    3129    3130    3131    3146    3147    3148    3149    3150    3151    3152    3153    3154    3155    3156    3157
1                           A       B       C       D                                               X       Y       Z

So what I am doing in my loop is:

df_main.merge(df_i, how='inner', on=df_i.columns)

Where, when i=1:

df_main.columns:

Index(['ID', '3120', '3121', '3122', '3123', '3124', '3125', '3126',
       '3127', '3128', '3129', '3130', '3131', '3146', '3147', '3148', '3149',
       '3150', '3151', '3152', '3153', '3154', '3155', '3156', '3157'],
      dtype='object')


df_i.columns:

Index(['3123', '3124', '3125', '3126', '3127'], dtype='object')

And code is raising this KeyError:

    raise KeyError(key)
KeyError: Index(['3123', '3124', '3125', '3126', '3127'], dtype='object')

How is this possible? df_i.columns is contained and exist in df_main.columns

Thank you in advance!

Upvotes: 0

Views: 339

Answers (1)

iamklaus
iamklaus

Reputation: 3770

okay one way to do this

df1

  3123 3124 3125 3126  3127
0    A    B    C    D   NaN

df2

  3146 3147 3148  3149  3150
0    X    Y    Z   NaN   NaN

using pd.concat

df = pd.concat([df.drop(df1.columns.append(df2.columns),axis=1),df2,df3], sort=True, axis=1)
df = df[['ID', 3120, 3121, 3122, 3123, 3124, 3125, 3126, 3127, 3128, 3129, 3130, 3131, 3146, 3147, 3148, 3149, 3150, 3151, 3152, 3153, 3154, 3155, 3156, 3157]] # for reordering
df.fillna('', inplace=True)

Output

    ID 3120 3121 3122 3123 3124 3125 3126 3127 3128 ...  3148 3149 3150 3151  \
0                      A    B    C    D           ...     Z                  

  3152 3153 3154 3155 3156 3157  
0                                

[1 rows x 25 columns]

Upvotes: 1

Related Questions