Sterling G. Baird comments

Results 550 comments of


                                            Sterling G. Baird

Change title of plot returned by `plot_marginal_effects` from "% better than experiment average" to "% higher..."

Haven't made a reproducer, but took a look at the functions: https://github.com/facebook/Ax/blob/57ba8714902ac218eb87dc2f90090678aa307a43/ax/plot/marginal_effects.py#L18-L69 From what I can tell, `greater_is_better` or equivalent isn't handled directly (just assumed) in `plot_marginal_effects`, and the only...

Change title of plot returned by `plot_marginal_effects` from "% better than experiment average" to "% higher..."

Here's an example of `plot_data.as_dict()` (using a breakpoint): ```python {'in_sample': {'0_0': PlotInSampleArm(name='0_0', parameters={'mu1_div_mu3': 1.5989093706011772, 'mu2_div_mu3': 46.885764732956886, 'std1_div_mu3': 85.80620403224603, 'std2_div_mu3': 6.035224910732358, 'std3_div_mu3': 93.63810631372034, 'comp1': 0.5191824799403548, 'comp2': 0.18046203907579184}, y={'vol_frac': 0.371}, y_hat={'vol_frac': 0.3727232216781629},...

Change title of plot returned by `plot_marginal_effects` from "% better than experiment average" to "% higher..."

Of course, the easy workaround in the interrim is to update the yaxis title (see https://github.com/facebook/Ax/issues/726#issuecomment-1095927782 for `to_plotly`): ## setup ```python import plotly.graph_objects as go from ax.plot.marginal_effects import plot_marginal_effects def...

Reparameterizing compositional linear equality constraints as linear inequality constraints introduces Sobol sampling bias

xref: https://github.com/facebook/Ax/issues/786

Reparameterizing compositional linear equality constraints as linear inequality constraints introduces Sobol sampling bias

@Balandat This is really helpful, thank you! [`sample_simplex`](https://github.com/pytorch/botorch/blob/ab04dd39a2d4c7734e41c5f26eb2dbba5b0e1771/botorch/utils/sampling.py#L270-L309) looks like it will help. I think one clarification is that I'm trying to sample from a simplex embedded in one higher...

Ideas for other models that could be "Lit"-ed

I will probably give `LitCrabNet` a go at some point. If I run into trouble, I will open another issue. Thanks!

[Information] Running Chalk on Windows

```bash (parables) PS C:\Users\sterg\Documents\GitHub\sgbaird\parables> npm run local > local > bundle exec jekyll serve --drafts C:/Ruby31-x64/lib/ruby/gems/3.1.0/gems/kramdown-1.17.0/lib/kramdown/parser/html.rb:10:in `require': cannot load such file -- rexml/parsers/baseparser (LoadError) from C:/Ruby31-x64/lib/ruby/gems/3.1.0/gems/kramdown-1.17.0/lib/kramdown/parser/html.rb:10:in `' from C:/Ruby31-x64/lib/ruby/gems/3.1.0/gems/kramdown-1.17.0/lib/kramdown/parser/kramdown/html.rb:10:in `require'...

[Information] Running Chalk on Windows

Ah😅case in point I guess. Got a little carried away with exploring multiple options and somehow missed/forgot this. Thanks @wobes1 !

Calculate metrics for quantile predictions

Also interested in this functionality https://github.com/ml-evs/modnet-matbench/issues/18

Convert from symmetric confidence intervals to standard deviation

https://handbook-5-1.cochrane.org/chapter_7/7_7_3_2_obtaining_standard_deviations_from_standard_errors_and.htm > If the sample size is large (say bigger than 100 in each group), the 95% confidence interval is 3.92 standard errors wide (3.92 = 2 × 1.96). So...