The first part of the address matching test should be: !is.na(name.df$address) (! was missing). Also instead of going for an exact match in the second part of the test I suggest using the Jarowinkler distance with a high similarity threshold in order to match up addresses, that differ only in details, maybe a value higher than 0.9 is advisable.
The first part of the address matching test should be: !is.na(name.df$address) (! was missing). Also instead of going for an exact match in the second part of the test I suggest using the Jarowinkler distance with a high similarity threshold in order to match up addresses, that differ only in details, maybe a value higher than 0.9 is advisable.