Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
  • Sep 09 14:18
    weikang9009 closed #175
  • Sep 09 03:59
    romass12 commented #175
  • Sep 07 14:27
    weikang9009 commented #175
  • Sep 07 14:27
    weikang9009 commented #175
  • Sep 07 14:26
    weikang9009 commented #175
  • Sep 07 06:47
    romass12 commented #175
  • Sep 07 03:43
    romass12 commented #175
  • Sep 06 15:31
    weikang9009 commented #175
  • Sep 06 05:52
    romass12 edited #175
  • Sep 06 05:51
    romass12 edited #175
  • Sep 06 05:51
    romass12 edited #175
  • Sep 06 05:51
    romass12 edited #175
  • Sep 06 05:51
    romass12 opened #175
  • Apr 01 18:07
    sjsrey synchronize #171
  • Apr 01 17:58

    sjsrey on master

    blacken repo update CI envs -36 +39,310 update unittests worklfow and 9 more (compare)

  • Apr 01 17:58
    sjsrey closed #173
  • Apr 01 17:58
    sjsrey commented #173
  • Apr 01 17:47
    sjsrey synchronize #171
  • Apr 01 17:43
    sjsrey synchronize #171
  • Mar 23 08:50
    martinfleis commented #173
Sergio Rey
@sjsrey
+1
Wei Kang
@weikang9009
Hi Serge, I need to leave now. I'll continue tomorrow.
Sergio Rey
@sjsrey
+1
Sergio Rey
@sjsrey
@weikang9009 if you are working on the notebook I can do one on the directional module as I've been working on some ideas. So don't worry about that one.
Wei Kang
@weikang9009
OK.
Wei Kang
@weikang9009
@sjsrey I created a pull request of adding notebooks for illustrating Markov and mobility methods.
Sergio Rey
@sjsrey
just merged it and am exploring it. looks great so far.
Wei Kang
@weikang9009
These two notebooks apply these methods to US data for demonstration purpose.
I plan to apply the rank-based measures to Mexico dataset (the one used in your 2016 AAAG paper), but the libpysal does not have the shapefile. I found one on the Internet and thinking of submitting a pull request for adding it into the example datasets. Does it sound reasonable to you?
Wei Kang
@weikang9009
The reason for not using US data is that intra- and inter-regional decomposition of rank-based measures requires a block weight matrix (or a regional scheme), and it would be quite convenient to use Mexican data as you investigated the impacts of regionalization scheme on inequality in your 2010 paper. These schemes are currently included in the example data set.
Sergio Rey
@sjsrey
An alternative would be to add a regimes variable to the us data set, since we already have the shapefile. BEA or Census regions?
or just add a variable with these definitions https://www.bea.gov/regional/docs/regions.cfm
Very nice job on the two notebooks.
Wei Kang
@weikang9009
OK! I'll create a notebook for applying rank based measures to US data. Can we also keep one for the application to Mexico as I have made a notebook for it?
Thanks!
Sergio Rey
@sjsrey
On mexico, there should be a shapefile in my repos for the paper: https://github.com/sjsrey/limaaag
Wei Kang
@weikang9009
Looks great! Do you think it would be good to include it in the example datasets of libpysal?
Sergio Rey
@sjsrey
sure
Wei Kang
@weikang9009
Just submitted a pull request for adding the Mexico shapefile to libpysal.
Wei Kang
@weikang9009
The notebook of applying rank measures to Mexico data is living in my giddy repo. I'll also start to construct another one using US data.
Wei Kang
@weikang9009
Some issues about the rank-based measures:
  • Should I add the code for local tau, LIMA and inter- and intra-regional decomposition of mobility association?
  • rank.py includes a class Theta , the regime mobility measure. Reference for this measure is "Rey, S. J. (2004). Spatial dependence in the evolution of regional income distributions. In Getis, A., Múr, J., and Zoeller, H., editors, Spatial econometrics and spatial statistics, pages 193–213. Palgrave, Hampshire.", but I could not find it in the paper. Could you please refer me to the right paper?
Sergio Rey
@sjsrey
Yes on the LIMA/tau code. And for the reference on Theta, see equation (8) here
Wei Kang
@weikang9009
OK!
Sergio Rey
@sjsrey
I think the pysal/giddy#16 now conflicts with your corrections, but you can see the fix for links and just add them to a new pr (you can close but not merge my conflicting PR)
Wei Kang
@weikang9009
Right. I'll submit a pr shortly.
Sergio Rey
@sjsrey
also note there were other changes in the README beside the links
Wei Kang
@weikang9009
I can see the "Note" on pypi and some format issue. Is that all?
Sergio Rey
@sjsrey
yup
Wei Kang
@weikang9009
Just submitted a pr to update README.md
Sergio Rey
@sjsrey
hey Wei i’m seeing errors in giddy when i test: as components.py is using pysal. did you change this?
Wei Kang
@weikang9009
Do you refer to line 1080 in markov.py?
I think we might have missed it and the unit tests did not catch it as there is not one for the method "spillover". The error was caught by doctest. I've corrected in the newly submitted PR.
Sergio Rey
@sjsrey
I fixed it, just wanted to make sure you didn’t have changes before.
Wei Kang
@weikang9009
I've included the change in the PR. Should I revert it?
Sergio Rey
@sjsrey
just close (don’t merge the pr)
hang on. i have one error left in markov.py due to using pysal.region
is that the one you fixed
Wei Kang
@weikang9009
Yes. I think so.
Sergio Rey
@sjsrey
ok. I’ll just copy your changes over to what i have from today and make sure all the tests pass. just leave your pr open for now.
Wei Kang
@weikang9009
OK!
Wei Kang
@weikang9009
Very slow right now!
Sergio Rey
@sjsrey
Pr is being updated for the description
Wei Kang
@weikang9009

There seems to be a problem with the notebook "Rose". When I try to open it locally, the following message shows:

Unreadable Notebook: /Users/weikang/Dropbox (ASU)/python_repos/pysal-refactor/giddy/notebooks/Rose.ipynb NotJSONError('Notebook does not appear to be JSON: u\'{\\n "cells": [\\n {\\n "cell_type": "...',)

It also won't be loaded in your github repo

Sergio Rey
@sjsrey
Rose can be deleted as the functionality is in the directional notebook
Wei Kang
@weikang9009
OK
Wei Kang
@weikang9009
Hi Serge, I was trying to test the release of giddy on the testing site, but was constantly running into the error:
Collecting esda (from giddy) Could not find a version that satisfies the requirement esda (from giddy) (from versions: ) No matching distribution found for esda (from giddy)
though the other dependency mapclassify can be successfully installed.
I wonder whether you had this issue when you released esda (libpysal is a dependency and I imagine the situation should be similar).
Sergio Rey
@sjsrey
@weikang9009 when you tested the installer for giddy did you test on python 2 and python 3?
Wei Kang
@weikang9009
I only tested on python 2. Sorry. Is there a problem on python 3?