List of Figures

• 2.1. The idealized integrated urban modelling system envisioned by Miller, Kriger & Hunt [34].
• 2.2. The link between list-based and contingency table representations of multivariate categorical data.
• 2.3. An illustration of the Iterative Proportional Fitting procedure with two variables $X$ and $Y$.
• 2.4. A simple example algorithm for the Iterative Proportional Fitting procedure using a two-way table and one-way target margins.
• 2.5. A zone-by-zone application of IPF for population synthesis, including a Monte Carlo integerization stage.
• 2.6. An illustration of Beckman et al.'s fitting procedure using two attributes $X$ and $Y$, plus a variable $Z$ representing the census tract zone within the PUMA.
• 3.1. The major groups within the Canadian census' person universe.
• 3.2. A breakdown of the Canadian census' person universe, by family membership.
• 3.3. A mosaic plot showing the breakdown of the SC86B01 summary tables: population by sex, age and highest level of schooling.
• 4.1. Simplifying the PUMS by removing high-dimensional associations.
• 4.2. A top-down algorithm for synthesizing persons and husband-wife families.
• 4.3. A bottom-up algorithm for synthesizing persons and husband-wife families.
• 5.1. Overview of complete synthesis procedure.
• 5.2. Diagram of the relationships synthesized between agents and objects, using the Unified Modelling Language (UML) notation
• 5.3. Algorithm showing conditional Monte Carlo synthesis using a sparse list-based data structure.
• 5.4. Map showing a dwelling attribute from the synthesized population.