cnliao / separation-explorer

Separation Explorer

A repository which contains the code to generate and to display the NIST ISODB separation explorer.

Data creation

All data scraping, sorting and processing performed can be found in the Jupyter notebooks located in the ./notebooks/ folder.

• scrape-isotherms: downloads the latest version of the NIST database and stores it in a single pickled file.
• select-isotherms: convert NIST isotherms to a pyGAPS format, followed by filtering and consolidation in an SQLite database.
• process-isotherms: generation of KPI and isotherm files for the explorer.
• compare-isotherms: comparisons between the NIST adsorption database and the MADIREL data.

Notebooks should be run with Python 3.6+

Data

In the ./data/ folder the following are found:

• iso.db - the complete ISODB database in a SQLite format
• iso-madirel.db - the complete MADIREL database in a SQLite format
• kpi.h5 - calculated KPI DataFrame, in a HDF5 format
• iso-packed.bak, .dat, .dir - simple shelve dictionary to store NIST isotherms

Dashboard

The separation explorer dashboard is built on top of a Bokeh server. The main code is found in ./src while auxiliary html templates, css and javascript can be found in the ./templates/ folder.

The dashboard is running on a Heroku dyno at https://pauliacomi.com/separation-explorer. Performance of this version is limited to the infrastructure and bandwidth provided by Heroku.

It is entirely possible (and faster) to run a local version of the explorer by first cloning this repository, installing all the Python requirements in requirements.txt and then starting a local Bokeh server with:

bokeh serve . --show