Skip to content

Latest commit

 

History

History
122 lines (93 loc) · 4.51 KB

README.md

File metadata and controls

122 lines (93 loc) · 4.51 KB
Raw

Scientific Benchmarking Results Visualizer - Barplot

This D3 graph is used to visualize the results of a benchmarking challenge that uses one single evaluation metric in the form of a Barplot. Challenge participants are shown in the X axis, while the value of their metric is shown in the Y axis.

figure1

Major Update v1.2

With v1.2 the package stops consuming an API, and starts consuming JSON structures, to improve reusability. (see below: Input)

NPM Package

NPM Package @inb/oeb-chart-barplot published to: https://www.npmjs.com/package/@inb/oeb-chart-barplot

Input

The visualizer uses as input the results of one challenge stored in the OpenEBench database in array format of the format of the official Benchmarking Data Model.

The necessary data structure for this visualization type is provided by the OpenEBench API https://openebench.bsc.es/api/scientific/widget/bar-plot/OEBD003000002S The array needs to be supplied in the data attribute data-data="[...]" with the following format:

data:   [
            {
                "toolname": "Ensemble_1",
                "metric_value": "0.562",
            },
            {
                "toolname": "Ensemble_2",
                "metric_value": "0.556",
            },
        ]

Furtheremore, the y axis name is set with the data-metric-name attribute, and the id with :data-id="ID". (Full example below)

Classification

As other OpenEBench results visualizers, this plot format results can be transfomed to tabular format by sorting the participants in descending/ascending order according to their metrics and applying a quartile classification over that lineal set of values. This classifcation splits the participants in four different groups/clusters depending on their performance. Clusters are separated in the plot with vertical lines and showed in the right table together with a green color-scale, which is easier to interpret for both experts and non-experts users.

figure2

Live Demo

See a demo of how this visualizer works here

How to use

The component can be imported in two way: As npm package (preferred), or via the build file from the git repository (see bottom).

Use the npm package

npm i @inb/oeb-chart-barplot

In your frontend component: import { load_bars_visualization } from "@inb/oeb-chart-barplot";

<div data-id="ID" data-data="[...]" data-metric-name="Metric Name" class="benchmarkingChart_bars" ></div>

You can then call the load_bars_visualization() function.

Attributes that can be set on the <div> tag

  • class: should always be 'benchmarkingChart_bars'
  • data-id : the official OEB id of the benchmarking challenge you want to visualize
  • data-data: Should always contain the metrics array as in https://openebench.bsc.es/api/scientific/widget/bar-plot/OEBD003000002S
  • data-metric-name: Should always contain the metric name

Example: <div data-id="OEBD003000002S" class="benchmarkingChart_bars" data-data="[...]" data-metric-name="..."></div>

Alternative way: Clone from repository

Requirements:

-npm -http server

Clone the repo to your document root :

git clone https://github.com/inab/Scientific_Barplot.git

Set within the index.html 'benchmarkingChart_bars' div, the OEB id of the dataset (data-id) you want to visualize:

<!DOCTYPE html>
<html lang="en">
    <head>
        <meta charset="UTF-8">
        <meta name="viewport" content="width=device-width, initial-scale=1.0">
        <meta http-equiv="X-UA-Compatible" content="ie=edge">
        <title>Scientific Benchmarking Barplot</title>
    </head>

    <body>
        <div
            class="benchmarkingChart_bars"
            data-id="OEBD003000002S"
            data-data=
            '[
                {
                    "toolname": "Ensemble_1",
                    "metric_value": "0.562",
                },
                {
                    "toolname": "Ensemble_2",
                    "metric_value": "0.556",
                },
            ]'
            data-metric-name="Metric Name"></div>
    </body>

</html>

Install dependencies from package.json :

npm install

Export node moodules :

export PATH="${PWD}/node_modules/.bin/:$PATH"

Compile with webpack and visualize sample results in your localhost :

./node_modules/.bin/webpack-cli src/app.js --output=build/build.js -d -w