Data and code for "Stochastic Modelling of Urban Structure" manuscript To run any of the Python files, potential_functions.c must be compiled into a shared object called potential_functions.so in the same folder. e.g. on OSX this can be done from the terminal with: gcc -fPIC -shared -o potential.so potential.c -O3 Examples were run with Python 3.6.0. with recent versions of numpy, scipy, matplotlib, ctypes, joblib and multiprocessing libraries installed. The following files relate to figures in the manuscript: - hmc.py HMC code to generate data for figure 5. - laplace_grid.py Likelihood values for figure 4. - mcmc_high_noise.py: MCMC scheme for gamma=10000 (low-noise regime) to generate data for figures 9-10. - mcmc_low_noise.py: MCMC scheme for gamma=100 (high-noise regime) to generate data for figures 7-8. - opt.py: Optimization routine to generate data for figure 6. - potential_2d.py: Illustration of 2d potential function to produce figure 2. - read.py: Plot MCMC and optimisation data saved down into the output directory. Will produce similar to figures 3, 5, 6, 8, 10 and stats relating to figures 7 and 9. - rsquared_analysis.py: R-squared analysis for deterministic model as discussed around figure 4. - data/london_n/ Directory containing datasets for the case study. Residential data is residential.csv and retail data is small_london.txt. Remaining files are pre-processed versions for simulations.