Downloading templates

Use our pre-configured templates as a starting point for your projects.

Selected templates

Tip

Click on the template link to see the source code on GitHub. Once there, you’ll see an option to launch a free, hosted JupyterLab.

  • Exploratory Data Analysis
    1. Basic EDA example: Load and clean data. Then create HTML reports with visualizations. (templates/exploratory-analysis)

  • Machine Learning
    1. Basic ML example: Get data, clean it, and train a model. (templates/ml-basic)

    2. Intermediate ML example: Create training and batch serving pipelines. (templates/ml-intermediate)

    3. Online API: Deploy pipeline as an API using Flask. (templates/ml-online)

    4. Experiment grid + Mlflow: Create a grid of experiments and track them with MLflow. (templates/mlflow)

  • SQL databases
    1. Basic SQL example: Process data, dump it, and visualize with Python. (templates/spec-api-sql)

    2. ETL: dump data from remote storage, upload it to a database, process it, and visualize it with Python. (templates/etl)

Downloading a template

To download any of the examples:

ploomber examples -n {template} -o {output-directory}

For example, if you want to copy the Basic EDA example to the eda directory in your computer:

ploomber examples -n templates/exploratory-analysis -o eda

Tip

Once you download an example, you can explore it with Jupyter. Check out the Jupyter integration guide to learn more.

Once the download finishes, you’ll need to install dependencies before you can run the template; you can use the ploomber install command, which automatically figures out if it should use conda or pip. If you prefer so, you may call conda or pip directly.

Warning

To plot the pipeline (ploomber plot command), you must install pygraphviz.

Listing all templates

To list all the available examples:

ploomber examples

Note that the command above will display three sections:

  1. Templates. Pre-configured projects that you can use as a starting point.

  2. Cookbook. Short examples to get something done quickly.

  3. Guides. In-depth tutorials covering features in detail.

Note that both Cookbook and Guides are part of the documentation itself, and you can navigate to any of them using the left sidebar or download them to run them locally.

Templates structure

All templates follow the same structure:

  1. README.md: Instructions to run the template.

  2. README.ipynb: Same as README.md but in notebook format and with command outputs.

  3. environment.yml: conda dependencies file.

  4. requirements.txt: pip dependencies file.

Most templates contain a pipeline.yaml file, so you can run ploomber build to execute the pipeline, but there are a few exceptions. Check out the template’s README.md for specifics.

Starting projects from scratch

If no template suits your needs, use the ploomber scaffold command to create a barebones project. Click here to learn how to scaffold projects.

ploomber scaffold also comes with utilities to modify existing pipelines, so can use it to modify any of the templates.