Montgomery County, MD Crime Predictor Project

Dash app

In development. Going over unfinished work and will be deployed sometime soon.

Installation

Fork the master branch, then clone the master branch to your local machine

>> git clone https://github.com/<repo>

Create virtual environment and install requirements.txt

>> pipenv install requirements.txt

NOTE: OPTIONAL

If you did the previous step and created an environment for this project, open the env

>> pipenv shell

Run the Dash app locally and start building!

run it by executing the following line

>> python run.py

Additionally, run the app locally before pushing with gunicorn

>> gunicorn run:server

Now, follow along Ryan Herr’s guide and build as you will.

Data

• The data source can be found at dataMontgomery

• The cleaned data can be found in the notebooks folder, named Cleaned_Crime.csv

• The smaller size cleaned data can be found in the notebooks folder, named sml_crime.csv

Notebooks

All of the data exploration can be found there.

• For the cleaning process, refer to Montgomery_Cleaning.ipynb

• For the visual creation process, refer to Montgomery_Visualization.ipynb

• For the Model Process, refer to Montgomery_Model.ipynb

The basis for this project has to do with wanting to find out how my intuition matched with the data on crime occurance in Montgomery County, MD; being where I lived for the majority of my childhood.

This project was our Unit2 Final Project at Lambda.

Model

• The joblib file can be found here finalized_model.joblib

• The explainer shap model can be found here explainer.joblib

Medium Article

Find the short read @yonipineda1010/predicting-type-of-crime-in-montgomery-county-md-26a4b375948">here