This is super cool. I always wanted a programming language that would track error bounds for me. So when I write $F = m*a$ and I have an $m = 100\pm 1kg$ and and $a = 9\pm 1m/s^2$ it would track this in calculations and let me extract the propagated bounds like $F.max$. I had only uniform and Gauss distributions in minds but allowing arbitrary distributions is of course way more powerful. Also less efficient but I guess many common cases could be optimized.
Have you looked into other probabilistic programming languages? While Squiggle could help in the use case you are describing, Stan and PyMC3 may be usable for the case you are describing.
Squiggle would treat that similar to Guesstimate, though also make it possible to write simple functions.
This is super cool. I always wanted a programming language that would track error bounds for me. So when I write $F = m*a$ and I have an $m = 100\pm 1kg$ and and $a = 9\pm 1m/s^2$ it would track this in calculations and let me extract the propagated bounds like $F.max$. I had only uniform and Gauss distributions in minds but allowing arbitrary distributions is of course way more powerful. Also less efficient but I guess many common cases could be optimized.
Thanks!
Have you looked into other probabilistic programming languages? While Squiggle could help in the use case you are describing, Stan and PyMC3 may be usable for the case you are describing.
Squiggle would treat that similar to Guesstimate, though also make it possible to write simple functions.