Mercurial > repos > goeckslab > tabular_learner

comparison README.md @ 1:f69ed50c9768 draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
planemo upload for repository https://github.com/goeckslab/gleam commit 4dc221b2fa9717552787f0985ad3fc3df4460158
author goeckslab
date Sat, 21 Jun 2025 15:06:53 +0000
parents 209b663a4f62
children
comparison
equal deleted inserted replaced
0:209b663a4f62 1:f69ed50c9768
1 # Galaxy-Pycaret 1 # Tabular Learner Tools
2 A library of Galaxy machine learning tools based on PyCaret — part of the Galaxy ML2 tools, aiming to provide simple, powerful, and robust machine learning capabilities for Galaxy users.
3 2
4 # Install Galaxy-Pycaret into Galaxy 3 This repository contains two machine learning tools for working with tabular data in the Gleam framework:
5 4
6 * Update `tool_conf.xml` to include Galaxy-Pycaret tools. See [documentation](https://docs.galaxyproject.org/en/master/admin/tool_panel.html) for more details. This is an example: 5 ## 1. Tabular Learner
7 ```
8 <section id="pycaret" name="Pycaret Applications">
9 <tool file="galaxy-pycaret/tools/pycaret_train.xml" />
10 </section>
11 ```
12 6
13 * Configure the `job_conf.yml` under `lib/galaxy/config/sample` to enable the docker for the environment you want the Ludwig related job running in. This is an example: 7 A comprehensive tool for training and evaluating multiple machine learning models on tabular datasets.
14 ```
15 execution:
16 default: local
17 environments:
18 local:
19 runner: local
20 docker_enabled: true
21 ```
22 If you are using an older version of Galaxy, then `job_conf.xml` would be something you want to configure instead of `job_conf.yml`. Then you would want to configure destination instead of execution and environment.
23 See [documentation](https://docs.galaxyproject.org/en/master/admin/jobs.html#running-jobs-in-containers) for job_conf configuration.
24 * If you haven’t set `sanitize_all_html: false` in `galaxy.yml`, please set it to False to enable our HTML report functionality.
25 * Should be good to go.
26 8
27 # Make contributions 9 ### Features:
10 - Supports both classification and regression tasks
11 - Automatically compares multiple algorithms to find the best model
12 - Extensive customization options:
13 - Data normalization
14 - Feature selection
15 - Cross-validation
16 - Outlier removal
17 - Multicollinearity handling
18 - Polynomial feature generation
19 - Class imbalance correction
20 - Outputs detailed HTML reports with performance metrics and visualizations
21 - Saves the best model for later use
28 22
29 ## Getting Started 23 ## 2. PyCaret Predictor/Evaluator
30 24
31 To get started, you’ll need to fork the repository, clone it locally, and create a new branch for your contributions. 25 A companion tool for making predictions and evaluating trained models on new data.
32 26
33 1. **Fork the Repository**: Click the "Fork" button at the top right of this page. 27 ### Features:
34 2. **Clone the Fork**: 28 - Works with models trained by Tabular Learner
35 ```bash 29 - Supports both classification and regression tasks
36 git clone https://github.com/<your-username>/Galaxy-Pycaret.git 30 - Generates predictions on new data
37 cd <your-repo> 31 - Creates evaluation reports when target values are provided
38 ``` 32 - Outputs predictions in CSV format
39 3. **Create a Feature/hotfix/bugfix Branch**:
40 ```bash
41 git checkout -b feature/<feature-branch-name>
42 ```
43 or
44 ```bash
45 git checkout -b hotfix/<hoxfix-branch-name>
46 ```
47 or
48 ```bash
49 git checkout -b bugfix/<bugfix-branch-name>
50 ```
51 33
52 ## How We Manage the Repo 34 ## Workflow
53 35
54 We follow a structured branching and merging strategy to ensure code quality and stability. 36 These tools are designed to work together:
37 1. Use **Tabular Learner** to train and find the best model for your dataset
38 2. Use **PyCaret Predictor/Evaluator** to apply your trained model to new data
55 39
56 1. **Main Branches**: 40 Both tools are powered by [PyCaret](https://pycaret.org/), an open-source machine learning library that automates the ML workflow.
57 - **`main`**: Contains production-ready code.
58 - **`dev`**: Contains code that is ready for the next release.
59
60 2. **Supporting Branches**:
61 - **Feature Branches**: Created from `dev` for new features.
62 - **Bugfix Branches**: Created from `dev` for bug fixes.
63 - **Release Branches**: Created from `dev` when preparing a new release.
64 - **Hotfix Branches**: Created from `main` for critical fixes in production.
65
66 ### Workflow
67
68 - **Feature Development**:
69 - Branch from `dev`.
70 - Work on your feature.
71 - Submit a Pull Request (PR) to `dev`.
72 - **Hotfixes**:
73 - Branch from `main`.
74 - Fix the issue.
75 - Merge back into both `main` and `dev`.
76
77 ## Contribution Guidelines
78
79 We welcome contributions of all kinds. To make contributions easy and effective, please follow these guidelines:
80
81 1. **Create an Issue**: Before starting work on a major change, create an issue to discuss it.
82 2. **Fork and Branch**: Fork the repo and create a feature branch.
83 3. **Write Tests**: Ensure your changes are well-tested if applicable.
84 4. **Code Style**: Follow the project’s coding conventions.
85 5. **Commit Messages**: Write clear and concise commit messages.
86 6. **Pull Request**: Submit a PR to the `dev` branch. Ensure your PR description is clear and includes the issue number.
87
88 ### Submitting a Pull Request
89
90 1. **Push your Branch**:
91 ```bash
92 git push origin feature/<feature-branch-name>
93 ```
94 2. **Open a Pull Request**:
95 - Navigate to the original repository where you created your fork.
96 - Click on the "New Pull Request" button.
97 - Select `dev` as the base branch and your feature branch as the compare branch.
98 - Fill in the PR template with details about your changes.
99
100 3. **Rebase or Merge `dev` into Your Feature Branch**:
101 - Before submitting your PR or when `dev` has been updated, rebase or merge `dev` into your feature branch to ensure your branch is up to date:
102
103 4. **Resolve Conflicts**:
104 - If there are any conflicts during the rebase or merge, Git will pause and allow you to resolve the conflicts.
105
106 5. **Review Process**: Your PR will be reviewed by a team member. Please address any feedback and update your PR as needed.