-
I frequently find myself wanting to include various bits of existing deterministic Javascript code in my WebPPL programs: for example, THREE.js, or some miscellaneous utility functions only relevant t…
-
- x Is refactor
- \_ Is new feature
- \_ Concerns documentation
# Description of suggestion or shortcoming:
WebPPL has this:
https://webppl.readthedocs.io/en/master/functions/other.html?highlig…
-
Example:
~~~~
>> Error: [{name: globalStore, loc: null, type: Identifier, comments: null, typeAnnotation: null}, {"name":"_k32","loc":null,"type":"Identifier","comments":null,"typeAnnotation":null…
-
-
We should implement streams in WebPPL. More generally, the functional reactive programming model would be pretty natural and useful to adopt. If our data were a stream we could do something like `data…
-
Comments on [webPPL documentation](http://docs.webppl.org/en/master/), to be taken with a grain of salt!
### Quick start
- Maybe the very first bullet should be, "If you haven't yet, install nodejs"?
…
-
I use webppl pretty regularly currently to learn how to do bayesian inference, and I notice that some distributions are missing that are in other offerings like pyMC3. One particular one is the Stude…
-
sweet.js gives us reasonable error messages for JS syntax errors. Errors for invalid webppl code which does parse as JS are not so nice. For example:
``` js
var x = /foo/
```
... results in:
```
{ …
-
I've [added callbacks](https://github.com/probmods/webppl/pull/312) to webppl's `MCMC` method will allow us to experiment with an inference progress interface at some point.
As a proof of concept, I …
-
For complex programs, it is useful to bundle up functions into objects. For example, we might want to represent agents as follows:
``` js
var agent = {
act: function(state){...},
expectedUtilit…