I have two pandas df with the following structure. Let's say df_1 is:
| Index | Column_A | Column_B |
|---|---|---|
| Index_1 | 4 | 8 |
| Index_2 | 7 | 1 |
| Index_3 | 5 | 9 |
| Index_4 | 4 | 8 |
| Index_5 | 2 | 3 |
and df_2 is:
| Index | Column_C | Column_D |
|---|---|---|
| Index_1 | 11 | 25 |
| Index_4 | 23 | 16 |
| Index_5 | 12 | 42 |
I want to concatenate the columns in the corresponding indices, so that the final df would be:
| Index | Column_A | Column_B | Column_C | Column_D |
|---|---|---|---|---|
| Index_1 | 4 | 8 | 11 | 25 |
| Index_2 | 7 | 1 | NaN | NaN |
| Index_3 | 5 | 9 | NaN | NaN |
| Index_4 | 4 | 8 | 23 | 16 |
| Index_5 | 2 | 3 | 12 | 42 |
But doing the following:
df = pd.concat([df1, df2], axis=1)
Outputs the error:
InvalidIndexError: Reindexing only valid with uniquely valued Index objects
Is there a way to make pandas ignore the indices of df2 not present in df1 and do the concatenation asigning NaNs to them? BTW in my real problem I have lots of dfs to concatenate, so it would be fantastic to have some option and just put the list of dfs to concatenate, while preserving the indices of the first one:
df = pd.concat([df1, df2, df3...], axis=1)
