The sample dataset

In order to make it easier to learn and practice Envision, we provide the following two sample datasets. These dataset are not real data, but we have made significant efforts to make sure it is similar to the data that can be found in a real-world supply chain. All the code snippets provided in the following are expected to work when applied to these dataset.

Dataset overview

The dataset represents the historical data of a small fictitious retailer. The dataset includes multiple tabular files formatted as flat text files.

The sample dataset contains the following files:

This dataset includes many interesting pieces of information such as product categories, stock levels, selling and purchasing prices, lead times, etc. It is also representative of the data that can typically be obtained from a company that uses a not-too-ancient ERP.

Using the dataset

You can use the links above to download the TSV files. Then, go to the Files tab of your Lokad account, create a folder named /sample – the path /sample is the convention that we follow in the rest of the documentation, and re-upload these files into this newly created folder. Your Lokad account should look like this:


Dataset details

In this section, we document all the fields contained in the sample datasets’ different files.


This file contains the list of products sold by the retailer along with relevant attributes at the product level.


This file contains the historical sales data disaggregated down to the transaction level, as the client identifiers are also provided.


This file contains the historical purchase orders data disaggregated down to the transaction level of every shipment.


This file contains the list of suppliers with their respective properties. On purpose, this file is not indexed by an Id column but by a Supplier column.


This file contains the list of bundles with their respective composition. The purpose of this file is to illustrate how Envision can cope with bills of materials.