Skip to content

Latest commit

 

History

History
226 lines (140 loc) · 10.9 KB

README.md

File metadata and controls

226 lines (140 loc) · 10.9 KB

Escalation

What is Escalation?

Escalation is a web app that runs a lightweight visualization dashboard for data analysis. You can set up plots and tables that update along with your data or analysis, and have interactivity options to allow for data browsing.

Some user cases for Escalation:

  • A research group that wants to have better visibility into experiments conducted by its members, or an at-a-glance record of experimental progress
  • A project team at a small company that wants to share progress or up to date results with management

All that is required for is to set a config file, either by hand or using the included UI Configuration Wizard, and you can run Escalation locally or on a server for others to view.

We have a demo app running and here.

Wait, aren't there already lots of different visualization tools out there?

Yes. Escalation has a few advantages:

  • Straightforward and low-cost deployment
  • Improved data privacy: everything can be hosted and run locally or on a server you control
  • Open-source code
  • Integration with data versioning and analysis pipelines (In development)

Limitations of Escalation

  • Currently uses a SQL database to store data. Escalation provides functionality to automatically create a SQL database from CSV files

Setup

What do you need for the app to work?

Note, because of Docker issues with Windows, there may be substantial hiccups running on that operating system. The instructions below have been tested on Mac and Linux. Linux users will need sudo permissions to run the following setup commands.

  • Configure the app
    • Escalation uses configuration files (json) to build the dashboard organizational structure, link the data in visualizations, and construct the visualizations themselves.
    • These configuration files can be built by hand, using the Configuration Wizard, or any combination of the two
  • Data
    • Escalation includes tooling to ingest CSV files into SQL, automatically building the necessary SQL data tables and the code necessary to integrate them with Escalation.
    • ToDo: Data Migration helpers- what happens when the format of your data changes over time?
  • Python environment to run the app
    • You need a Python environment set up to run the web app. See instructions for setting up an environment, using Docker to handle the environment for you.

Each of these components are discussed further below.

1. Set up your local environment to run the web app and database using Docker

From the root level of the code repository, run:

docker-compose build

(this takes a little while the first time, as components are downloaded and installed)

We recognize that Docker is less common in academic settings, but highly recommend using it. Here are instructions on getting started using Docker. Additionally, we use Docker Compose for running our multi-container application, which does depend on the Docker Engine. Once you have Docker successfully installed, your next step is to install Docker Compose. Here are instructions for the installation process. We use the Docker containers to run our configuration wizard, as well as the scripts to ingest csv data into a SQL database. Once we set up a configuration and your data, we'll also use these containers to run the web app.

2. Configuring the app

2.1 Run the Configuration Wizard

Run the configuration wizard app from the root directory of this repo:

./escalation/scripts/wizard_launcher.sh

This launches the UI Configuration Wizard in a Docker container. Navigate in your the web app in your browser at: http://localhost:8000/wizard or http://127.0.0.1:8000/wizard This app runs in debug mode, and should detect the changes you make as you edit the configuration. Refresh your browser to update the contents to match your saved configuration.

Some notes on creating your first config files with the UI wizard.

2.2 Load your data

SQL database backend

Escalation provides functionality to parse csv data files, determine the relevant sql schema, create tables in the SQL database from your file, and create the necessary models.py file for the app to interact with the database. This creates sql tables that can be used by the graphics and tables on your Escalation dashboard.

In the configuration wizard in your browser, navigate to http://localhost:8000/upload

Use this web form to upload each file you'd like to use for your visualizations and include in the database. Note, it may take a little while to run.

If you'd like to add more than one csv to the same table, you have two options: Load them at the same time (by holding shift and selecting multiple files), or submit the additional CSVs using the "Append to existing table" option.

Todo: If you have an existing SQL database, how do you copy it into Escalation? We require each table to have an upload_id column, and a key that is unique within an upload_id value (i.e, the pair (upload_id, row_index) is a unique key.)

Todo: Add csvs programatically using a script

3. Running the configured app

We recommend running the Escalation web app using the docker container:

Re-run the docker compose build command to re-launch the containers with the app including all of the configuration you just did:

docker-compose up --build -d

To use the app, navigate in your browser to: http://localhost:8000 or http://127.0.0.1:8000

To shut down the app and stop the containers:

docker-compose down  

Interacting with the SQL database

The app creates a SQL database running in a Docker container which stores uploaded data. In addition to using the Escalation web app functionality, you can interact with this data as you would any other SQL database, writing queries or sql subroutines.

To connect directly to the SQL database, this command

docker exec -it escos_db psql -h localhost -p 5432 -U escalation -d escalation

Resetting the SQL database

Run the reset script from the top level of the repository

./escalation/scripts/escalation-reset.sh

This deletes the volume associated with the database, where the data is stored, and resets a models.py code file that decribes the database.

Simply re-start the configuration as above(#2.-Configuring-the-app) to work from this blank state.

You can also manually delete tables in the sql database by connecting to the database directly and using sql commands. Note that this doesn't reset the models.py file, so may result in strange behavior.

Connect to db:
docker exec -it escos_db psql -h localhost -p 5432 -U escalation -d escalation

List tables:
\dt

Drop table:
DROP TABLE my_table;

Disconnect:
\q

Uploading data files to the db from the command line

You can directly upload files by POSTing requests to the Escalalation upload endpoints.

http://localhost:8000/wizard/upload http://localhost:8000/upload

There is a helper script in escalation/scripts/csv_to_sql.py that does this, with documentation there about how to use it.

Running Escalation as a web-accessible server

In summary, you'll build a Docker-compose image that includes your dashboard configuration, and deploy it on a web-accessible server. This can be a server on a local network, e.g. in your lab, or on a cloud provider like DigitalOcean, Heroku, or AWS.

Escalation-specific instructions for Digital Ocean here.

Todo: Heroku, AWS

Building your Docker image for deployment

  1. You'll want to change the default settings for the password and username for the database, defined here: escos/escalation/app_deploy_data/app_settings.py
  2. Todo: instructions to build docker image and push

How can I contribute? (advanced)

Running Locally, without using Docker (useful for Escalation code development, testing)

You can also set up a custom Python virtual environment and run the server locally as you would any other Flask web app. ToDo: More detailed instructions on virtual env setup, requirements install, and running the app. Include info about db connection from host to Docker db

Developing for Escalation

  • optional, but recommended: run in a virtual environment of your choice. If you use conda and not pyenv or similar, use the appropriate package installer in the next step
  • pip install -r requirements-dev.txt
  • pre-commit install sets up the pre-commit hooks to auto-format the code. This is optional, the repo is formatted with Flake and Black.
  • From the escalation/ directory (at the same level as app,py, run export FLASK_ENV=development && python -m flask run

How to add a new type of plot

Development for Escalation has focused on Plotly, Cytoscape, and Seaborn, but the codebase should be compatible with other libraries or custom graphics. If you want to use something other than the provided libraries, your code should:

  • Inherit from graphic_class.py
  • Be added to available_graphics.py
  • Include an html file with javascript code required to plot
  • In addition, if you would like to use the wizard with the new graphic, you should define a schema with a class that inherits from graphic_schema.py. The schema should follow JSON Schema.

How to add a new option feature

  • add it to available_selectors.py
  • create a html document input elements need name "<id>|<type>|<column_name>"
  • add to create_data_subselect_info, modify_graphic_dict_based_on_addendum_dict, and add_operations_to_the_data_from_addendum as appropriate in graphic_class.py or one of its inheritors.
  • build in functionality in graphic_class or data_storer inheritor

Using and citing Escalation

Code submitted to Zenodo for DOI registration and version tracking. Bibtex and similar citation formatting available here:

DOI

License

The Apache 2.0 License applies to all code and materials associated with Escalation.

Copyright [2020] [Two Six Labs, LLC]

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.