Do some travel purposes lead to more tourist expenditure patterns than others? Evidence from an almost ideal demand system (AIDS) analysis