skale-me / skale-examples

[DEPRECATED] A place to share skale sample applications

skale-examples

Description

A place to share skale sample applications

Requirements

We assume you have already installed Node.js and skale to run those examples.

Install

First clone this repository to your local computer:

git clone https://github.com/skale-me/skale-examples.git

Then navigate to one of the example folders and install dependencies

cd skale-examples/wordcount
npm install

Note: For some examples you may need additional tools in your environment, like gnuplot for the adult application.

Finally, run the sample app

skale run # To run it on the skale cloud
skale test # If you have skale-engine running locally

Examples

wordcount

Compute the number of occurences of each word of James Joyce's Ulysses.

pi

Compute pi by throwing darts on the unit circle. For more information on the algorithm see https://en.wikipedia.org/wiki/Approximations_of_%CF%80#Summing_a_circle.27s_area

adult

This example shows machine learning capabilities of skale-ml and skale-engine.

We load and featurize a training dataset and a validation dataset from files.

We standardize features to zero mean unit variance variables.

We train a binary logistic regression model using a Stochastic Gradient Descent.

We generate the Receiver Operating Characteristic curve as a png image.

skale-examples's Issues

Update the examples to use a node based charting module

One barrier to making the examples work is that a few of them rely on gnuplot, which isn't the easiest to install outside of linux. I suggest that the examples be updated to create the plot images using a node-based charting module, such as:

https://d3js.org/
http://www.chartjs.org/
http://gionkunz.github.io/chartist-js/
http://dygraphs.com/

All of those are browser based as far as I know, but it might be possible to export to a png. D3 mentions this in their docs:

https://github.com/d3/d3/wiki#supported-environments