I use Haskell, but for many data manipulation problems it is still clearer to do them in Excel, mostly because of the bigger emphasis on the data itself, easier visualisation and eyeball-verification intermediate results, and simpler data-garbage-fixing in case a data point "raises an exception".