https://github.com/stanfordnlp/plot-interface
Web interface for the plotting project
https://github.com/stanfordnlp/plot-interface
Last synced: 3 months ago
JSON representation
Web interface for the plotting project
- Host: GitHub
- URL: https://github.com/stanfordnlp/plot-interface
- Owner: stanfordnlp
- Created: 2017-06-02T02:12:11.000Z (over 8 years ago)
- Default Branch: master
- Last Pushed: 2018-11-29T15:41:40.000Z (about 7 years ago)
- Last Synced: 2025-08-31T03:04:19.632Z (4 months ago)
- Language: JavaScript
- Size: 95.2 MB
- Stars: 3
- Watchers: 10
- Forks: 2
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# plot-interface
user interface for plotting with semantic parsing
## Running locally
* Clone the repository
* install yarn and run `yarn install`
* start client that queries the local server: `yarn start:local`
* start client that queries the public server `yarn start`
* push a version to the public url: `yarn run build && yarn run deploy`
Note that if you run locally on one machine,
it will not work to visit that machine from any other machine.
To make this work, you can use `REACT_APP_SEMPRE_SERVER=http://:8405 yarn start`
## Turking
* send turking parameters through url parameters: `BASE_URL/?uid=workerId&sid=assignmentId` Make sure the parameters are AFTER the # in the url so it appears in the very end. like http://localhost:3000/#/build?uid=blah
* this should be logged in the server at `SEMPRE_ROOT/plot-out/query.jsonl` by default
check that the parameters are loaded correctly!
## Picking examples
* data for vegalite and example specifications are in `public`
* `public/spec/vega-lite/index.json` together with `src/helper/vega-utils.responsesFromExamples` determines which examples get displayed in the beginning.
* To download examples `svn export https://github.com/vega/vega-lite.git/trunk/examples/specs public/spec/vega-lite`
* To get data use `svn export https://github.com/vega/vega-datasets.git/tags/v1.19.0/data/ public/data`
## Labeling procedure
Label JSON can be used to manually label some data, the sempre readable example will be printed to the javascript console.
## Workflow
see [schema documentation](https://github.com/vega/vega-loader#data-format-parsing)