emdash

The goal of emdash is to provide a deployer dashboard for an e-mission deployment.

The dashboard is structured as follow:

Dashboard
- Value boxes
  - Number of users (+ daily)
  - Number of active users today
  - Number of trips (+ daily)
  - Time since first user
- Plots
  - Sign ups
  - Trips
  - Branch
  - Platform

Tables
- Interactive plot (linked with the tables)
- Tables (downloadable in various formats from the UI)
  - Participants
  - Trips

Maps
- Trips (with a data filter on the left side)

Installation

You can install emdash from GitHub:

install.packages("remotes")
remotes::install_github("asiripanich/emdash")

How to use the dashboard

Run locally

To use the dashboard you must config the port of your e-mission mongoDB in the emdash::run_app(mongo_url = 'mongodb://localhost:27017') command. As you can see, the default is 27017. The easiest way to use this dashboard with your e-mission deployment is to create a SSH tunnel to the e-mission mongo on your e-mission server.

ssh -L {my-local-port}:localhost:{remote-server-port} username@your-server-ip-address-or-domain-name
ssh -L 27017:localhost:27017 ec2-user@123.12.12.1

Here is how you launch a e-mission dashboard with emdash.

library(emdash)
emdash::run_app(mongo_url = 'mongodb://localhost:27017')

Loading test data

This repository contains a test mongodump with entries from the e-mission unit test real data.

2016-07-22
2016-07-25
2016-07-27
2016-08-04
2016-08-05
2016-08-09
2016-08-10
2016-08-11

It includes background sensed data (background/*), analysis results (analysis/*) and the mode and purpose confirm objects from the master branch (manual/mode_confirm and manual/purpose_confirm).

To load the data into a local mongodb,

$ unzip test-data/2016_jul_aug_test_trips.zip
$ mongorestore

If you would like to load additional data, you can do so through the e-mission server.

For convenience, you may use this docker_start_test_mongodb.sh for starting up a Docker container that will run the test database.

Use with Docker

We provide a Dockerfile to run this dashboard. In the future we may include this as part of an e-mission docker-compose file which will be hosted on the e-mission-docker github repo.

Be warned that the emdash Docker image is quite large, around 2.8 GB, and take a while to build.

How to customise the dashboard

First you need to know R and Shiny. I tried to modularise each component as much as possible so that the dashboard is easy to be customised. Any advices to improve the dashboard are welcomed. :)

Customise query functions

See R/utils_query_database.R.

The functions in this script are for querying data from the e-mission mongoDB database.

Customise data

See R/utils_tidy_data.R.

The functions in this script are for cleansing and transforming data queried using the functions in R/utils_query_database.R.

ahcyip/emdash