clabrugere / ecommerce-dash

Example of a Dash application applied to e-commerce data to represent key metrics in the form of an interactive dashboard

Sales Dash application

This is a small project showing capabilities of Plotly Dash. It consists in a simple example of an e-commerce dashboard to monitor sales with some business KPIs and visualization based on transactional data with a simple ARMA sales 15 days look-ahead forecast. It is based on the public data Brazilian E-Commerce Public Dataset by Olist available on Kaggle.

The dataset contains around 100k transactional data about orders on a Brazilan e-commerce platform named Olist. Details on the dataset can be found on Kaggle.

Demo

A live demo of the app can be found here

Getting Started

Just clone this repository, navigate to the directory and execute the command:

python app.py

Requirements

• Python 3.7
• numpy
• pandas
• dash
• plotly
• statsmodels

License

This project is licensed under the MIT License - see the LICENSE.md file for details

Acknowledgments

• Olist and André Sionek, “Brazilian E-Commerce Public Dataset by Olist.” Kaggle, doi: 10.34740/KAGGLE/DSV/195341.
• Dash Natural Gas Well Production demo from which I largely took inspiration.

ecommerce-dash's Issues

Issue with Arima model

Hi, i took your app as a reference for one my ecommerce app.Kindly use below module to fix the raise error in ARIMA prediction.

from statsmodels.tsa.arima_model import ARIMA is got depreciated.

I could see new error after updating module to "from statsmodels.tsa.arima.model import ARIMA"
server: <Flask 'app'>
loading data: <dash.dash.Dash object at 0x000001D5B69611F0>
computing forecast data:
File "E:\My_Poc\reporting\ecommerce-dash\src\model.py", line 13, in predict
model_fit = model.fit(
File "E:\My_Poc\reporting\ecommerce-dash\venv\lib\site-packages\statsmodels\tsa\arima\model.py", line 318, in fit
self._spec_arima.validate_estimator(method)
File "E:\My_Poc\reporting\ecommerce-dash\venv\lib\site-packages\statsmodels\tsa\arima\specification.py", line 731, in validate_estimator
raise ValueError('"%s" is not a valid estimator.' % estimator)
ValueError: "css" is not a valid estimator.