We consider domestic hot water (DHW) consumption hourly data for Finnish apartments in November and August. Using datasets obtained in a previous work, we formulate a bottom-up model to quantify correlations in the consumption patterns, which are discerned by a different number of occupants for both weekday (WD) and weekend (WE). The analytical formulas thus obtained describe accurately the hourly consumption of any specific dataset. In particular, we can generate the consumption curves for unknown datasets and derive quantitatively the correlations between occupant groups and different seasons. We explain this procedure into details, define the key variables of the model and validate it against the measurements. Our quantitative results are immediately applicable to simulation tools for energy investigations and sizing of heating systems in Finland or areas with similar occupant behavior. More generally, the analytical, inductive method here introduced could be adapted to DHW studies concerning other geographic areas as well. We also argue that this simple, yet effective formalism might also be extended to other engineering contexts that are not strictly related to energy consumption. For example, the main idea could be developed and adapted to those disciplines where understanding dataset correlations constitutes an important investigation tool.