v3.7.0 - Total recoding, UX feats, and combined errors fix
Relevant changes on v3.7.0:
- Recode: got rid of data.table dependency for r2py wrapper and removed all
globalVariables
associated noise - Recode: all code is now clean and formatted under the tidyverse code style for better code reading and standardization
- Feat: trimmed functionality for response curves on one-pagers outputs to have coherent ranges plotted
- Feat: enabled channels removal on
robyn_allocator()
by setting their constraints to 0 #411 - Feat: when manually selecting refresh models in
robyn_refresh()
, re-ask user until valid solID is provided, instead of crashing - Feat: new
plot
and improvedprint
methods forrobyn_refresh()
outputs - Feat: include time units used in adstock plots for clarity
- Feat: enabled organic media variables to be calibrated (no spend)
- Fix: when best model based on minimum combined errors was tied with other models, inconsistent outputs (one-pagers / clustering). Standardized combined errors methodology with new
errors_scores()
function, especially normalizing errors before filtering models. The largest the "error_score", the better the model's performance #428 - Fix: show blue dots on top of grey dots in Pareto plots #420
- Fix: positive/negative colour palette on waterfall plot when all values are positive
- Fix: set prophet's print as disabled when prophet_vars input is NULL (off)
- Docs: added CRAN, site, and FB group badges on README files
- Docs: several typos and documentation updates
Full Changelog: v3.6.3...v3.7.0