sayari-analytics/pyisic

Full concordance graph

JSPromisel opened this issue · 4 comments

JSPromisel commented

Description

Given the API outlined, having a separate object for each standard1 -> standard2 concordance seems to make sense. But I imagine there might be some utility to having a full graph with all standards, which would allow you to do something like, show me all concordance across all standards for a given code in a given standard. Not to mention we could probably make a cool graph viz for the docs.

Environment

  • OS
  • Python version
  • Package version

Details

If necessary, describe the problem you have been experiencing in more detail.

Code To Reproduce

If reporting a defect, please include a code sample to reproduce the defect and outline the expected behavior.

b-say commented

cool graph viz using Trellis?

  • 🚀1
jvani commented

Agreed that it could be neat and could make for an interesting visualization.

That said, I have only been adding concordances to go to ISIC4, if we want to add reverse concordances (they're out there and might be useful for others), then a global graph would be a multi-digraph with cycles.

JSPromisel commented

So are we making the assumption that a code in standardA mapped to a code in standardB does not imply the inverse (e.g. code in standardB -> code in standardA)? Or am I misinterpreting concordance here?

jvani commented

So are we making the assumption that a code in standardA mapped to a code in standardB does not imply the inverse (e.g. code in standardB -> code in standardA)? Or am I misinterpreting concordance here?

Yes, this is not necessarily true. See the concordance tables here: https://unstats.un.org/unsd/classifications/Econ