Monte Carlo Simulation Effeciency #115
Description
https://github.com/npr99/intersect-community-data/blob/main/ncoda_07hv4_HUA_PREC_NSI.ipynb
New version provides a loop to step through the process by updating the random seed. Each new iteration will have slightly different results.
For Seaside (Clatsop County) which has about 40,000 people. 10 iterations took about 15 minutes and produced 3.2 GB of data.
One recommendation for your workflow. After each iteration you can create a table that saves the random seed and the key statistics that you are interested in. For example, after each iteration I could save the percent dislocation across multiple subgroups (renters, owners, low income, middle income, high income etc). Once the statistics are saved you can delete the HUA files that are created with each iteration.
| seed | dislocation | Renter dislocation | Owner Dislocation | Low Income Renter Dislocation |
|---|---|---|---|---|
| 9876 | 0.20 | 0.25 | 0.19 | 0.28 |
| 9877 | 0.21 | 0.24 | 0.20 | 0.27 |
ISSUE: Work on cleaning up the number of files made after the first iteration. Provide functions to save results in a table and then delete files.