David Kirkby
David Kirkby
I modified my code so that each thread build its own SBProfile objects, but things just got worse. Now `SersicInfo::kValue()` is throwing a `Table has repeated arguments.` exception at this...
I modified `LRUCache::get` to always return a new value, thus completely disabling the SersicInfo cache, which finally gives the same results with different numbers of threads. This is not a...
This looks like a nice approach. The method we use in the DESC [WeakLensingDeblending package](http://weaklensingdeblending.readthedocs.io/en/latest/) is to read the LSST simulated catalog ("CatSim"). Working directly with the catalog brings in...
KDE allows you to smooth and interpolate, but not extrapolate, right?
Has there been any recent progress in defining a formal grammar for the docopt DSL that can be used to automatically generate parser code in different languages?
I have the same issue, using the docker instructions. An additional problem with the docker instructions is that I needed to: ``` export PYTHONPATH=$pwd ``` in order to find the...
http://www.r2d3.us/visual-intro-to-machine-learning-part-1/ has great visualizations of the general ML setup and concepts and decision trees in particular.
Simpson's paradox https://pwacker.com/simpson.html
Humorous video related to gradient descent (or, equally, grad student desecent): https://www.youtube.com/watch?v=cvuktushEhY
Review these lecture notes on ML for Physicists: https://sites.krieger.jhu.edu/jared-kaplan/files/2019/04/ContemporaryMLforPhysicists.pdf