A final project made by Ellie Jensen and Niels Vanderloo for Bob Bosch's MATH 331 (Linear Optimization) in Fall 2022.

A linear program implemented with Gurobi to solve for the optimal city pairs for high speed rail in the United States. See included PDF of the paper optimizing_trains.pdf describing the model we implement.

• Python
• Jupyter notebooks
• the Gurobi Python package
  • installed with python -m pip install gurobipy

To run the analysis, first run the cells of the combinedStat.ipynb notebook in order. Then run the optimize.py file.

• minDist is the number of miles in our travel demand calculation where travel demand no longer improves as $d < $minDist
  • This shows up in our model as $T_{i, j}=\left(\frac{P_i^{0.8} \times P_j^{0.8}}{\max \left(d_{i, j}, {\tt minDist}\right)^2}\right)$

• trackAmt is the "budget" you have for building track
• penaltyAmt is the amount you assign as a penalty in track miles to decrease your track "budget" for each new route you add
• perCity is the maximum number of train routes that any given city can have

• combStat.csv contains the latitude, longitude and Census population data of every combined statistical area in the united states
  • this is used in combinedStat.ipynb to generate the other csv's used in the optimize.py program