Clean up df after all ods import

[jwestw]

After importing all the ODS files and concatinating them, there are quite a lot of columns. We need to work out what to do with these.

In one case, the sampling_point was found to be the same data as retail_outlet, so when one was populated the other was empty. This has been fixed.

But there are other mostly empty column(s), like packer_/_manufacturer_/_importer which might be the same data as another column, like the previous example.

These columns need to be investigated and tidied up by combination or deletion.

• 'packer_/_manufacturer'
• 'product', 'packer_/manufacturer/_importer'
• 'address_postcode'
• 'packer__postcode'