{"id":20780742,"url":"https://github.com/farrajota/benchmark_mice_algorithms","last_synced_at":"2026-05-05T08:31:25.657Z","repository":{"id":202488453,"uuid":"154053935","full_name":"farrajota/benchmark_mice_algorithms","owner":"farrajota","description":"Benchmark of Multiple Imputation using Chained Equations (MICE) algorithms on missing value imputation","archived":false,"fork":false,"pushed_at":"2018-10-21T22:21:48.000Z","size":216,"stargazers_count":1,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"master","last_synced_at":"2025-09-12T04:08:58.214Z","etag":null,"topics":["algorithms","benchmark","docker","jupyter","jupyter-notebook","mice","notebook","python"],"latest_commit_sha":null,"homepage":"","language":"Jupyter Notebook","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/farrajota.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null}},"created_at":"2018-10-21T21:28:37.000Z","updated_at":"2021-06-05T07:49:39.000Z","dependencies_parsed_at":"2024-05-10T23:00:57.746Z","dependency_job_id":null,"html_url":"https://github.com/farrajota/benchmark_mice_algorithms","commit_stats":null,"previous_names":["farrajota/benchmark_mice_algorithms"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/farrajota/benchmark_mice_algorithms","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/farrajota%2Fbenchmark_mice_algorithms","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/farrajota%2Fbenchmark_mice_algorithms/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/farrajota%2Fbenchmark_mice_algorithms/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/farrajota%2Fbenchmark_mice_algorithms/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/farrajota","download_url":"https://codeload.github.com/farrajota/benchmark_mice_algorithms/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/farrajota%2Fbenchmark_mice_algorithms/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":32641989,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-05-04T10:08:07.713Z","status":"online","status_checked_at":"2026-05-05T02:00:06.033Z","response_time":54,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["algorithms","benchmark","docker","jupyter","jupyter-notebook","mice","notebook","python"],"created_at":"2024-11-17T13:39:01.637Z","updated_at":"2026-05-05T08:31:25.628Z","avatar_url":"https://github.com/farrajota.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Benchmark of Multiple Imputation using Chained Equations (MICE) algorithm\n\nThis repo contains a benchmark of the MICE algorithm regarding performance and execution time for imputing missing values on data.\nFor this purpose, several variations of the MICE algorithm have been implemented using [LightGBM](https://github.com/Microsoft/LightGBM) instead of linear models for value imputation for accuracy and speed improvements. Here, four method are evaluate against the mean / mode value imputation procedure for multi-dimensional data on 5 different datasets available on the **Scikit-Learn** package (namely, boston house prices, iris, diabetes, wine and breast cancer):\n- **Vanila MICE**: Value by value imputation over a set number of iterations\n- **Fast MICE**: Column by column imputation over a set number of iterations\n- **Slow-Fast MICE**: Value by value imputation in the first iteration and column by column for the remaining iterations\n- **Fast-Slow MICE**: Column by Column imputation in all iteration except the last one where value by value imputation is used for the remaining iterations\n\nThe procedure is available via a jupyter notebook in the `notebook/` folder in this repo.\n\n## TL;DR\n\nIf you are just looking for the results of the benchmark, here they are:\n\n- On average, Fast MICE is **12.0x** faster than Fast-Slow / Slow-Fast MICE and **56.0x** faster than Vanila MICE\n- On average, Fast-Slow / Slow-Fast MICE are **5.0x** faster than Vanila MICE\n\n### Boston house prices results\n\n\u003cp align=\"center\"\u003e\u003cimg src=\"imgs/benchmark_results_boston_plot.png\" alt=\"Network architecture\" height=\"100%\" width=\"100%\"\u003e\u003c/p\u003e\n\n### Iris results\n\n\u003cp align=\"center\"\u003e\u003cimg src=\"imgs/benchmark_results_iris_plot.png\" alt=\"Network architecture\" height=\"100%\" width=\"100%\"\u003e\u003c/p\u003e\n\n### Diabetes results\n\n\u003cp align=\"center\"\u003e\u003cimg src=\"imgs/benchmark_results_diabetes_plot.png\" alt=\"Network architecture\" height=\"100%\" width=\"100%\"\u003e\u003c/p\u003e\n\n### Wine results\n\n\u003cp align=\"center\"\u003e\u003cimg src=\"imgs/benchmark_results_wine_plot.png\" alt=\"Network architecture\" height=\"100%\" width=\"100%\"\u003e\u003c/p\u003e\n\n### Breast Cancer results\n\n\u003cp align=\"center\"\u003e\u003cimg src=\"imgs/benchmark_results_breast_cancer_plot.png\" alt=\"Network architecture\" height=\"100%\" width=\"100%\"\u003e\u003c/p\u003e\n\n## Requirements\n\n- Python3 (3.6 recommended)\n- jupyter\n- [scipy stack](https://www.scipy.org/stackspec.html) (pandas, scipy, scikit-learn, etc.)\n- docker (optional, recommended)\n\n## Getting started\n\nThe code is available via jupyter notebooks for easier use.\n\nTo run these notebooks, you need to start a jupyter server. Here, you can do it in two ways:\n\n- a) run a local jupyter server or\n- b) run a self-contained docker image.\n\n### Run a local jupyter server\n\nTo start the jupyter server you must first have python + jupyter installed. The quickest way to accomplish this is by installing [anaconda](https://www.anaconda.com/download/).\n\nAfter installing anaconda, you should create an environment:\n\n```bash\n$ conda create -n py36_jupyter python=3.6 anaconda\n```\n\nThis command will install the recommended version of CPython and the necessary packages to run the code.\n\nFinally, to start a jupyter server you simply need to run the following command:\n\n```bash\n$ jupyter notebook\n```\n\n### Run a self-contained docker image\n\nTo run the notebooks using docker, you first need to build the container's docker image. To do so, you just need to do the following:\n\n- i) Build the container using a Makefile macro:\n\n    ```bash\n    $ make build\n    ```\n\n- ii) Run the container using a command:\n\n    ```bash\n    $ docker image build -t jupyter_scipy_custom .\n    ```\n\nThen, to start the container you can:\n\n- i) Run the container using a Makefile macro:\n\n    ```bash\n    $ make run\n    ```\n\n- ii) Run the container using a command:\n\n    ```bash\n    $ docker run --rm -p 8888:8888 -v \"$PWD\"/notebook:/home/jovyan/work --name jupyter_benchmark_mice jupyter_scipy_custom\n    ```\n\n## License\n\n[MIT](LICENSE)","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ffarrajota%2Fbenchmark_mice_algorithms","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Ffarrajota%2Fbenchmark_mice_algorithms","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Ffarrajota%2Fbenchmark_mice_algorithms/lists"}