Skip to content

Latest commit

 

History

History
152 lines (97 loc) · 4.09 KB

README.md

File metadata and controls

152 lines (97 loc) · 4.09 KB
Raw

Pivot

Pivot is a web-based exploratory visualization UI for Druid built on top of Plywood.

The project is currently undergoing rapid development. Internal and external APIs may change with little notice.

Features

Drag-and-drop UI

Drag to Split

Contextual exploration

Time Highlight

Comparisons

Time Highlight

Usage

Ensure that you have an up-to-date node

Make sure you have node (>= 4.x.x) installed. On MacOS with homebrew you can do:

brew update
brew install node

Install

Next simply run:

npm i -g imply-pivot

That's it. You are ready to Pivot.

Example

Start off by running an example (static) dataset:

pivot --example wiki

Run with Druid

Next connect Pivot to your broker by simply pointing it to your broker host

pivot --druid your.druid.broker.host:8082

Pivot will automatically introspect your Druid cluster and figure out your dimensions and measures.

Note: if Pivot starts up and gives you a query error it is most likely because it could not properly introspect your schema. You probably have some hyperUnique column that Pivot is trying to SUM over. You will have to provide Pivot with a config file as in the nest section.

Create a config

In general Pivot will never know your schema as well as you. To get a better experience you should create a config and provide it to Pivot. The fastest way to create a config is to have Pivot do it for you.

pivot --druid your.druid.broker.host:8082 --print-config --with-comments > config.yaml

The --print-config option will make Pivot run through its regular introspection and then, instead of spinning up a server, dump the YAML onto the stdout and exit.

pivot --config config.yaml

The next step is to examine and tweak the config using your favorite editor nano config.yaml.

Development

Here are the steps to clone Pivot and run it as a developer.

Firstly make sure you have the latest node (>= 5.5.x) and gulp installed:

npm i -g gulp

Clone the project

git clone [email protected]:implydata/pivot.git
cd pivot

Inside the pivot folder run:

npm install
gulp

Finally you have to create a config.yaml file. (or use the sample)

./bin/pivot --druid your.druid.broker.host:8082 --print-config --with-comments > config.yaml

The --with-comments flag adds docs about what goes into the config. Alternatively you can read the comments in the sample config file.

Then you are ready to

./bin/pivot --config config.yaml

We use WebStorm 11 to develop Pivot and the checked in .idea directory contains all of the auto formatting and code styles. You are free to use any editor as all the build scripts are editor agnostic.

Running gulp watch will build the project and start all the automated watchers.

Roadmap

Recent improvements:

  • Relative filters
  • All new introspection code
  • Better comparison behavior and legend interaction
  • Support for query time lookups (ex. $language.lookup('wikipedia-language-lookup'))
  • Support for the extract function (ex. resourceVersion: $resource.extract("(\d+\.\d+\.\d+)"))

We will be working on:

  • Additional visualizations (bar chart, geo, heatmap)
  • Exclusion filters
  • String / RegExp filters
  • Removing strict limits on dimension values
  • Better time selection with date picker
  • Various additions, improvements and fixes to make the app more complete

Questions & Support

For updates about new and upcoming features follow @implydata on Twitter.

Please file bugs and feature requests by opening and issue on GitHub and direct all questions to our user groups.