I have a DataFrame with non-unique column names. What I would like to do is append either a series or a single column dataframe to a given column within the existing dataframe on the index based on that DataFrame column index location (as opposed to the column name, which would overwrite values in columns with the same name). Basically I am looking for something that works like pd.df.columns[1].append.
Asked
Active
Viewed 250 times
0
riddley_w
- 249
- 1
- 4
- 7
-
What should be the values in other columns after that? Nulls? How is it different from adding a row? http://stackoverflow.com/questions/10715965/add-one-row-in-a-pandas-dataframe – sashkello Feb 13 '14 at 23:26
-
The other columns would be nulls until they were updated. The larger context of what I am trying to do is update a timeseries indexed dataframe via json api calls, but each column needs to be a separate query. – riddley_w Feb 13 '14 at 23:41
-
So, why can't you append a row, as above and then fill it in with data as it comes? – sashkello Feb 13 '14 at 23:51
-
The intent is to be appending hundreds of rows at a time, and I think I would still be faced with the same issue of directing the data to the proper column – riddley_w Feb 13 '14 at 23:54