This project aims to analyze crime information in Tallahasse, FL, collected from TOPS.
This project contains two parts:
- Traditional data analysis on TOPS data using Geopandas and Scikit-learn
- Crime heatmap prediction given a geographical map, following He, Zheng (2021)
First, clone this repository:
git clone git@github.com:mao1756/tlhcrimemap.git
Then, you can install the required packages as follows:
pip install -r requirements.txt
It is assumed that you use Python 3.9 to run the script. (I believe this does not matter except for the data collection from ArcGIS - see details in the data collection script)
Please first see the project slides for the project summary.
On top of the dataset available in the dataset folder, you can also collect new data from TOPS using our data collection script. Please take a look at the comments in the script on how to use it.
In addition to the slides, see our EDA notebook for detailed analysis.
To train pix2pix, you first need to create the image dataset for geographical and heat maps. For this, you can use our processing script. After that, you need to use the combining script from the pix2pix repository. For the explanation of our script, see the comments in it, and for the script from pix2pix, see their document. After you have made the dataset, you can train pix2pix using their script. For details, see their documentation.