Gravar-mail: Models of epidemics: when contact repetition and clustering should be included