-
tfp.bijectors.IteratedSigmoidCentered simulates the stick breaking process. That means while using in forward or backward the output must be positive as they are probabilities.
When I run following …
-
@pkofod and I thought it would be useful to brainstorm about possible GSoC 2018 projects.
If anybody has ideas for projects or feedback on the points below that would be great.
**Constrained opti…
-
Using both the CRAN version and dev version of `ergm`, this fit:
```r
nw
-
~~This may relate to our issues involving optimisation of frames/orientations:~~
- When running multiple, supposedly independent runs we get memory corruption. I've been noticing this in both Pytho…
-
From official D-Wave docs:
> The D-Wave 2000Q QPU has up to 2048 qubits and 6016 couplers.
For solving the tracking problem, you converted it into the QUBO problem.
If you solve it on D-Wav…
-
When `scipy.optimize.curve_fit` is called with the `bounds` keyword, the input data is not checked for proper size. This results in weird error messages from the lapack library, when the input data is…
-
@burdges [had an idea](https://github.com/rust-lang-nursery/rand/pull/303#issuecomment-373539213), looking something like:
```rust
pub trait BlockRngCore {
/// Results element type, e.g. `u32`.…
-
#### Summary:
In several Dirichlet models where i'm modeling a a vector of probabilities in each observation, I get an error with boost library when I try to use `vb`. This is on Linux 18.04, rstan…
-
Mentioned by @ahwillia in the [JOSS review](https://github.com/openjournals/joss-reviews/issues/615#issuecomment-373460590).
> Is there a nice API to do a nested optimization? Let's say I wanted t…
-
Use interpolation (see Nocedal's book) to decrease the line-search step.