How to start

• You should be able to run the code with default conda environment. If not, try to create Python environment using gmgwr_env.yml following this guide.
• As the first trial, run mgwr/tests/main.py, see more at mgwr/tests/README.md.

Structure

• The core functions locate in mgwr
• Some testing scripts and sample results locates mgwr/tests

Benchmark

https://github.com/pysal/mgwr/

• equal_interval() in search.py has bug. when running MGWR with .search(search_method="interval"), l_bound and u_bound is not correctly passed and get error TypeError: unsupported operand type(s) for +: 'NoneType' and 'int'
• In this package, bw = 100 means a local regression with 100 neighboring points including itself. In Pro GWR, bw = 101 means 100 neighboring points + 1.

Reference for Trust Newton Algorithm

Julia

• http://julianlsolvers.github.io/Optim.jl/v0.9.3/algo/newton_trust_region/
• https://optimization.mccormick.northwestern.edu/index.php/Trust-region_methods

Trust Region Newton Method for Large-Scale Logistic Regression

• https://jmlr.csail.mit.edu/papers/volume9/lin08b/lin08b.pdf

NEWTON’S METHOD FOR LARGE BOUND-CONSTRAINED OPTIMIZATION PROBLEMS

• https://www.csie.ntu.edu.tw/~cjlin/papers/tron.pdf
• https://www.mcs.anl.gov/~anitescu/CLASSES/2012/LECTURES/S310-2012-lect5.pdf

saddle point

• https://people.orie.cornell.edu/dsd95/SIOPTsaddles.pdf
• https://arxiv.org/pdf/2006.01512v3.pdf
• https://github.com/hphuongdhsp/Q-Newton-method/blob/master/src/functionsDirectRun.py