c3js / c3js/c3

Gaussian Process Ribbon Style Uncertainty Area

Open
#1,923 0 comments 0 reactions 0 assignees View on GitHub
Dominant language
JavaScript
Stars
9.3k
Forks
1.4k
Avg merge
6d 16h
Merged PRs (30d)
1

Description

I have a function with some upper and lower bounds which express some uncertainty. The style of plot I would like to achieve looks like this

http://www.gaussianprocess.org/gpml/code/matlab/doc/f3.gif
or
http://gpytest2.readthedocs.io/en/latest/_images/tuto_GP_regression_m2.png

The closest I have been able to get to this using the following code from proto-repl-charts format

```
(proto-repl-charts.charts/custom-chart
"Custom"
{:data {:columns
[(cons "x" grid)
(cons "gp" (map (comp probit F) grid))
(cons "gp+" (map (fn [x] (probit (+ (F x) 0.5))) grid))
(cons "gp-" (map (fn [x] (probit (- (F x) 0.5))) grid))]
:types {:gp "spline" :gp+ "area" :gp- "area"}
:colors {:gp- "grey" :gp+ "grey"}
:x "x"}})
```

resulting in

![screenshot from 2017-01-06 00-02-47](https://cloud.githubusercontent.com/assets/4892846/21708198/06241a38-d3a4-11e6-9c41-11ab3f91de49.png)

This is already pretty good, but ideally I do not want the area below the lower curve to be grey, I want it to be white.

Does anyone have a better way of coding this using what is currently available? I searched the docs for "ribbon" but couldn't find anything. This would be a really great feature to add.

Contributor guide

Open the contributing guide

Research direction

Start with the proto-repl-charts custom-chart configuration and its area rendering options, then compare the referenced Gaussian-process ribbon examples with the current result. Done means the uncertainty band is shown between the upper and lower curves while the area below the lower curve remains white, using currently available functionality or a documented feature addition.

Written by the indexing model from the issue text.

Assessment

Tech stack
javascript
Domain
data-visualization
Issue type
Feature
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.