written by Eric J. Ma on 2019-10-31 | tags: data science deep learning testing pair programming code review
I share my thoughts on why re-implementing deep learning models _from scratch_ can be a very valuable activity.
Read on... (1603 words, approximately 9 minutes reading time)written by Eric J. Ma on 2019-10-30 | tags: essays data science workflow good practices
My thoughts on how to do code review in data science projects effectively.
Read on... (598 words, approximately 3 minutes reading time)written by Eric J. Ma on 2019-10-29 | tags: data science drug development artificial intelligence medicine
A mini-rant on why I think hyped-up "AI" has a very long way to go, and that medicine is much too complex for us to solve at one shot.
Read on... (531 words, approximately 3 minutes reading time)written by Eric J. Ma on 2019-10-18 | tags: python tips optimization packages
`cachier`, a really nifty tool for caching function results, is really useful and easy-to-use! Come read why.
Read on... (253 words, approximately 2 minutes reading time)written by Eric J. Ma on 2019-10-05 | tags: jupyter dataops devops data science
Some notes on how to serve up an HTTPS-enabled Jupyter server.
Read on... (425 words, approximately 3 minutes reading time)written by Eric J. Ma on 2019-10-05 | tags: teaching bayesian statistics
Do we use multiple examples to highlight the same point, or do we take one example and layer on complexity? It all depends.
Read on... (187 words, approximately 1 minute reading time)written by Eric J. Ma on 2019-09-07 | tags: data apps data science devops deployment
Some of the benefits of using Dokku as a PaaS solution for data scientists. It's free, open source, and when paired with the right tools, enables data scientists to focus on making their data app prototypes.
Read on... (794 words, approximately 4 minutes reading time)written by Eric J. Ma on 2019-07-29 | tags: software development sprint code sprint open source community community development
Post-PyCon, I detail some of my thoughts on how one can productively participate in a code sprint as a participant.
Read on... (872 words, approximately 5 minutes reading time)written by Eric J. Ma on 2019-07-26 | tags: data science data products app deployment
The key things I learned building my first Panel app: prototype in the notebook, use `.servable()` on the thing to serve up, test locally, and use Heroku!
Read on... (172 words, approximately 1 minute reading time)written by Eric J. Ma on 2019-07-23 | tags: data science statistics distributions
It’s mainly because they are the generalization of two distributions, the Cauchy and Gaussian. Come learn more!
Read on... (196 words, approximately 1 minute reading time)