overview of these notes
I have separated analysis into 2 parts:
- original analysis with paper, jupy, etc
- new analysis with python and jinja
little memory tricks
backups
The code is now effectively backed up within a git repository since it is in my code folder.
It is currently saved in 3 places:
- /home/chris-jakoolit/bayesian_dta_clone (this is the one that i work on after booting)
- cloudfare for bayesianpca.com
- /home/chris-jakoolit/christopherpaine_org/_code/bayesian_pca_DTR (ensures we have a backup but needs to be done manually with the following.)
rclone bisync /home/chris-jakoolit/christopherpaine_org/_code/bayesian_pca_DTR/ /home/chris-jakoolit/bayesian_dta_clone/ --progress
small to do list of items
- universal inference engines
- pymc3 library
- conjugate priors
some history on the workflow
im found jupyter notebooks the logic gets tangled with presentation so have separated concerns.