Table of Contents

1. Installation
2. Description
3. Data
4. File Descriptions
5. Results
6. Acknowledgements

Installation

• Python 3.7.2
• Libraries:
  • NumPy
  • Pandas
  • patsy
  • statsmodels
  • matplotlib
  • scikit-learn

You will also need to have software installed to run and execute an iPython Notebook

Description

The business problem was formulated as follows:

Pawdacity is a leading pet store chain in Wyoming with 13 stores throughout the state. This year, Pawdacity would like to expand and open a 14th store. Your manager has asked you to perform an analysis to recommend the city for Pawdacity’s newest store, based on predicted yearly sales.

Data

• p2-2010-pawdacity-monthly-sales.csv - This file contains all of the monthly sales for all Pawdacity stores for 2010.
• p2-partially-parsed-wy-web-scrape.csv - This is a partially parsed data file that can be used for population numbers.
• p2-wy-453910-naics-data.csv - NAICS data on the sales of all competitor stores where total sales is equal to 12 months of sales
• p2-wy-demographic-data.csv - This file contains demographic data for each city and county in Wyoming.

File Descriptions

You can find the results of the analysis in either html form or complete Jupyter Notebook:

• Pawdacity_New_Store_Location.html
• Pawdacity_New_Store_Location.ipynb

Alterinatively, run one the following commands in a terminal after navigating to the top-level project directory new_store_location (that contains this README):

ipython notebook Pawdacity_New_Store_Location.ipynb

or

jupyter notebook Pawdacity_New_Store_Location.ipynb

This will open the iPython Notebook software and project file in your browser.

Results

To identify the new city location based on potential sales, I performed the following steps:

• Step 1: Preprocessing

  • aggregated sales data from months to years
  • cleaned and joined census data with sales data and demographics data
  • dealt with outliers

• Step 2: Predictive Modeling

  • checked for corr() between predictor variables to avoid multicollinearity
  • removed certain features based on their variance_inflation_factor()
  • calculated OLS() Regression and removed not statistically significant predictors
  • predicted sales for new cities with the final linear regression model
  • filtered data according to provided criteria for the new city

Acknowledgements

Having started learning Python, I decided to rewrite the project I first completed in Alteryx within the Predictive Analytics for Business Nanodegree at Udacity.