James Brennan
A blog about statistics, data science, and remote sensing.
Home
All posts
Tags
22
Mar 2021
Building counterfactuals for sklearn models
Counterfactuals provide a model-agnostic method for black box machine learning algorithms interpretable and human-understandable.
12
Mar 2021
The Laplace approximation
The Laplace Laplace approximation still provides the simplest deterministic method available for approximate inference. Let's implement it in python.
06
Mar 2021
Horseshoe priors in numpyro
This post looks at how to implement a horseshoe prior in `numpyro` to do sparse Bayesian inference. We'll see that the ...
13
May 2020
Linking python with PostGIS
GIS databases are sort of a necessary evil. Once your data gets too big for RAM you've got to start thinking either abo...
11
Apr 2020
Estimating CDFs from data done right
This post looks at estimating empirical cumulative density functions (CDFs) and their confidence intervals from data. ...
23
Mar 2020
Confidence intervals for LOWESS models in python
LOWESS (or also referred to as LOESS for *locally-weighted scatterplot smoothing*) is a non-parametric regression metho...
16
Mar 2020
Fast and easy gridding of point data with geopandas
This post looks at using the geopandas library to do fast efficient gridding of point data onto a regular grid.
14
Mar 2020
Visualising 2 billion pixel rasters with datashader and xarray
This post looks at visualising large raster datasets with [Datashader](https://datashader.org) and [xarray](http://xarr...
06
Mar 2020
Essential packages for geospatial analysis in python
Python is fast becoming the defacto standard for geospatial work -- primarily because of its great ecosystem of package...
01
Mar 2020
ASCII maps of vector files
# ʕ•ᴥ•ʔ I recently came across a nice python package for visualising vector data quickly. Of course, an obvious option...