Restaurant Chain Sales Dashboard

Project Description

This is a development project for a restaurant KPI dashboard. Current features include:

• Generating simulated sales data for analytics pipeline development based on:
  • pre-programmed menu and location lists
  • regional and seasonal trends
  • day-of-week fluctuations
• Visualizing sales data in an interactive Dash web app

Technologies Used

• Python - version 3.11.4
• Pandas
• Plotly
• Pyarrow
• Dash

Features (Current)

• Interactive dashboard for visualizing sales data
• Data pipeline for generating synthetic sales data
• Data pipeline for converting CSV data to Parquet format using brotli compression for storage efficiency
• Dockerfile for containerizing the app

Feature additions (Roadmap)

• Each chart should be able to be clicked / highlighted, and when clicked, the underlying data table at the bottom should filter to show only the data that corresponds to the clicked chart

Dashboard Image

source data overview:

sample of source csv data before conversion to parquet:

region,location,date,category,menu_item,quantity_sold,net_sales
Southeast,Atlanta,2023-01-01,starters,Charcuterie Board,2,44
Southeast,Atlanta,2023-01-01,starters,Charcuterie Board,1,22
Southeast,Atlanta,2023-01-01,starters,Spinach & Artichoke Dip,2,18
Southeast,Atlanta,2023-01-01,starters,Calamari,5,60
Southeast,Atlanta,2023-01-01,starters,Spinach & Artichoke Dip,3,27
Southeast,Atlanta,2023-01-01,starters,Baked Brie,5,70
Southeast,Atlanta,2023-01-01,starters,Fried Pickles,5,40
Southeast,Atlanta,2023-01-01,starters,Baked Brie,5,70
Southeast,Atlanta,2023-01-01,starters,Spinach & Artichoke Dip,3,27
Southeast,Atlanta,2023-01-01,starters,Baked Brie,2,28

dataframe info:

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 7079822 entries, 0 to 7079821
Data columns (total 7 columns):
 #   Column         Dtype 
---  ------         ----- 
 0   region         object
 1   location       object
 2   date           object
 3   category       object
 4   menu_item      object
 5   quantity_sold  int64 
 6   net_sales      int64 
dtypes: int64(2), object(5)
memory usage: 378.1+ MB

date range:

• 2023-01-01 to 2023-12-31

total records:

• about 7,000,000

total locations:

• 15-20

granularity (each record / row has):

• total quantity_sold and net_sales for each menu_item per location per day for 2023

visualizations:

• Total net sales org-wide (a single metric at the top)
• 4 charts in a grid layout:
  • Total net sales per region
  • Total net sales per location
  • Total net sales per category
  • Top 25 menu items org-wide by total net sales (data table)

slicers:

• slicers for:
  • region
  • location
  • date range (start and end date)
  • menu category
  • menu item
• if no slicers are selected, the dashboard will show the full data

Local Installation & Setup (Mac)

Prerequisites:

• Python 3.11+

1. Clone the repository

git clone https://github.com/thefilesareinthecomputer/restaurant_sales_dashboard.git
cd restaurant_sales_dashboard

2. Set up a virtual environment

python3.11 -m venv {VENV_NAME}

source {VENV_NAME}/bin/activate

3. Install dependencies

pip install -r requirements.txt

4. Set up environment variables (optional)

• Create a .env file in the root of your project and define the following:

STAGING_DATA_DIRECTORY=/path/to/staging/data
APP_DATA_DIRECTORY=/path/to/app/data

5. Run the data pipeline (optional) (sample parquet file already included in the repo)

• If you want to edit the menu or locations lists, you can do so in the data_pipeline/restaurant_details.py file
• Then, to generate new synthetic data, run the following:

python data_pipeline/sales_data_creator.py

• This creates a CSV file in the STAGING_DATA_DIRECTORY defined in your .env file

python data_pipeline/sales_data_pipeline.py

• This converts the CSV file to a parquet file and saves it in the APP_DATA_DIRECTORY defined in your .env file

6. Run the app

python dash_app/app.py

7. Open the Dashboard

Visit http://127.0.0.1:8050/ in your browser.