On construction and (non)existence of c-(almost) perfect nonlinear functions