I have a PDF file containing tabular data spread across multiple pages. The first page of the PDF contains the column headers, while subsequent pages contain rows of data corresponding to those headers. When I use Python and tabula to extract the tables, I end up with a list of DataFrames, one for each page.

My goal is to combine these DataFrames into a single DataFrame, treating the entire PDF as a single table. However, when I attempt to concatenate the DataFrames using pd.concat, I encounter issues where rows with NaN values are introduced, likely due to the differing column structures between the first page (with headers) and subsequent pages (data rows only).

My code:
`#Reading pdf tables into a list of DataFrames
dfs = tabula.read_pdf(‘/content/drive/MyDrive/Python_dataset/APEX_Loans_Database_Table (3).pdf’, combined_df = pd.concat(dfs, ignore_index=True)

print(combined_df)
`

However, this approach results in rows with NaN values where the columns on subsequent pages don’t match the columns on the first page.

How can I concatenate these DataFrames into a single DataFrame while preserving the structure of the table and avoiding the introduction of NaN values?

Any insights or alternative approaches would be greatly appreciated. Thank you!