Totally a preference thing. I strongly prefer pandas to dplyr having worked with both.