Patrick Connolly patcon

The purpose of this document is to show how scikit-learn can be run in browser to fit PCA data and interactively transform arbitrary new samples from a user.

Usage

copy text of index.html below
paste it into any local directly (e.g., /tmp/index.html)
open it in a browser
- ```
 $ open /tmp/index.html
```

Source

Context: polis-community/red-dwarf#55

Apply with

curl https://gist.githubusercontent.com/patcon/d8b850c5e8c5bed0021b748cf224a051/raw/0001-Add-is-meta-back-into-comments.csv-export.patch
git apply 0001-Add-is-meta-back-into-comments.csv-export.patch

or

Except for the code segments that render the magnifying glass (as circle or path), this is essentially a copy of Mike Bostock's Fisheye Distortion example.

The motivation for this was largely aesthetical but also a bit functional. When using the Fisheye as an aid to viewing large datasets, I often find myself experimenting with the degree and radius of the distortion. In such cases, I find it helpful to see the exact reach of the distortion and then decide the best configuration for a particular dataset.

	<!DOCTYPE html>
	<html>
	<head>
	<style>
	#plot-wrapper {
	display: flex;
	width: 100%;
	gap: 20px;
	}
	svg {

	border: none
	height: 660
	license: CC0-1.0