# seglines
Compute segmented least squares for a dataset. Probably assumes that `X = 1...N` or something.
```
$ seglines 10 data.csv --plot
opt = 70.19
segment 1: 0 (1.330) 13 (13.880) f(x) = 1.009·x + 1.693
segment 2: 14 (1.340) 27 (14.190) f(x) = 0.997·x + -11.887
segment 3: 28 (3.350) 41 (15.760) f(x) = 0.987·x + -25.083
segment 4: 42 (3.420) 55 (17.630) f(x) = 0.991·x + -37.058
segment 5: 56 (5.160) 69 (18.600) f(x) = 0.949·x + -46.523
segment 6: 70 (3.870) 83 (17.660) f(x) = 0.973·x + -63.211
segment 7: 84 (5.930) 97 (17.490) f(x) = 0.977·x + -76.830
segment 8: 98 (5.210) 111 (17.590) f(x) = 0.897·x + -82.366
segment 9: 112 (3.810) 125 (16.410) f(x) = 0.963·x + -102.544
segment 10: 126 (16.780) 139 (3.190) f(x) = -1.036·x + 147.696
```
## Install
`pip install seglines`
Depends only on `numpy`. When using `--plot`, we also need `matplotlib`.
## Usage
There is a `--help` option: `seglines --help`
Use `seglines L data.csv` where `L` is the number of segments you want to segmentize into.
In case you want to generate an `L`-segmented linear dataset, use
`seglines --generate 5 10 > data.csv`
and then
`seglines 5 data.csv`
This will output the segments, e.g.
```
opt = 16.49
segment 1: 0 (0.410) 9 (10.330) f(x) = 0.961·x + 1.355
segment 2: 10 (3.750) 19 (13.260) f(x) = 0.987·x + -5.741
segment 3: 20 (13.530) 29 (4.210) f(x) = -1.031·x + 33.960
segment 4: 30 (13.880) 37 (7.420) f(x) = -0.913·x + 41.254
segment 5: 38 (5.190) 49 (14.720) f(x) = 0.904·x + -29.629
```
To create a plot of the dataset, add `--plot`:
`seglines 5 data.csv --plot`
Raw data
{
"_id": null,
"home_page": "https://github.com/pgdr/seglines",
"name": "seglines",
"maintainer": "PG Drange <Pal.Drange@uib.no>",
"docs_url": null,
"requires_python": "",
"maintainer_email": "",
"keywords": "seglines",
"author": "PG Drange",
"author_email": "Pal.Drange@uib.no",
"download_url": "https://files.pythonhosted.org/packages/99/2d/f6ef0fb44b2766d9e3a459705ccc1d44b29d3760eddd012719765b65d72c/seglines-0.0.5.tar.gz",
"platform": null,
"description": "# seglines\n\nCompute segmented least squares for a dataset. Probably assumes that `X = 1...N` or something.\n\n```\n$ seglines 10 data.csv --plot\nopt = 70.19\nsegment 1: 0 (1.330) 13 (13.880) f(x) = 1.009\u00b7x + 1.693\nsegment 2: 14 (1.340) 27 (14.190) f(x) = 0.997\u00b7x + -11.887\nsegment 3: 28 (3.350) 41 (15.760) f(x) = 0.987\u00b7x + -25.083\nsegment 4: 42 (3.420) 55 (17.630) f(x) = 0.991\u00b7x + -37.058\nsegment 5: 56 (5.160) 69 (18.600) f(x) = 0.949\u00b7x + -46.523\nsegment 6: 70 (3.870) 83 (17.660) f(x) = 0.973\u00b7x + -63.211\nsegment 7: 84 (5.930) 97 (17.490) f(x) = 0.977\u00b7x + -76.830\nsegment 8: 98 (5.210) 111 (17.590) f(x) = 0.897\u00b7x + -82.366\nsegment 9: 112 (3.810) 125 (16.410) f(x) = 0.963\u00b7x + -102.544\nsegment 10: 126 (16.780) 139 (3.190) f(x) = -1.036\u00b7x + 147.696\n```\n\n\n## Install\n\n`pip install seglines`\n\nDepends only on `numpy`. When using `--plot`, we also need `matplotlib`.\n\n\n## Usage\n\nThere is a `--help` option: `seglines --help`\n\nUse `seglines L data.csv` where `L` is the number of segments you want to segmentize into.\n\nIn case you want to generate an `L`-segmented linear dataset, use\n\n`seglines --generate 5 10 > data.csv`\n\nand then\n\n`seglines 5 data.csv`\n\nThis will output the segments, e.g.\n\n```\nopt = 16.49\nsegment 1: 0 (0.410) 9 (10.330) f(x) = 0.961\u00b7x + 1.355\nsegment 2: 10 (3.750) 19 (13.260) f(x) = 0.987\u00b7x + -5.741\nsegment 3: 20 (13.530) 29 (4.210) f(x) = -1.031\u00b7x + 33.960\nsegment 4: 30 (13.880) 37 (7.420) f(x) = -0.913\u00b7x + 41.254\nsegment 5: 38 (5.190) 49 (14.720) f(x) = 0.904\u00b7x + -29.629\n```\n\nTo create a plot of the dataset, add `--plot`:\n\n`seglines 5 data.csv --plot`\n\n\n",
"bugtrack_url": null,
"license": "MIT",
"summary": "Segmented Least Squares",
"version": "0.0.5",
"split_keywords": [
"seglines"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "992df6ef0fb44b2766d9e3a459705ccc1d44b29d3760eddd012719765b65d72c",
"md5": "3760f38a5c7293a26b47f2e7aeaece87",
"sha256": "54392c8598afd9c83f943b1d3dc3ce93e030e0585c6343002270ca6fb5c33fb4"
},
"downloads": -1,
"filename": "seglines-0.0.5.tar.gz",
"has_sig": false,
"md5_digest": "3760f38a5c7293a26b47f2e7aeaece87",
"packagetype": "sdist",
"python_version": "source",
"requires_python": null,
"size": 4824,
"upload_time": "2023-03-15T09:35:09",
"upload_time_iso_8601": "2023-03-15T09:35:09.486501Z",
"url": "https://files.pythonhosted.org/packages/99/2d/f6ef0fb44b2766d9e3a459705ccc1d44b29d3760eddd012719765b65d72c/seglines-0.0.5.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-03-15 09:35:09",
"github": true,
"gitlab": false,
"bitbucket": false,
"github_user": "pgdr",
"github_project": "seglines",
"travis_ci": false,
"coveralls": false,
"github_actions": false,
"lcname": "seglines"
}
JSON
