Kusti Skytén

Writing XMP metadata

2

I'm able to read XMP metadata with ```python pdf = PdfReader('example.pdf') metadata = pdf.Root.Metadata # {'/Length': '503', '/Subtype': '/XML', '/Type': '/Metadata'} xmp = metadata.stream ``` but I haven't quite figured...

I couldn't find documentation for any APIs of Polar. I'm looking for APIs for exporting all annotations for a specific file, generate new annotations and flashcards, and for bulk importing...

The Remarkable tablet is a popular e-reader with quite an active hacker community: https://github.com/reHackable/awesome-reMarkable. It would be really nice to be able to sync annotations both ways as long as...

Mounting search queries

3

Thanks for making tmsu, it's really neat. IIt would be really nice to be able to mount specific queries instead of having to go through the fixed structure of the...

Ducktype sampleable objects

8

The is a draft for adding arbitrary simulators [here](https://github.com/cscherrer/Soss.jl/blob/abc/src/simulation.jl). Would it be possible to just treat everything on the right hand side of a tilde as implementing a `rand` method...

Implement the TableTraits.jl interface

2

Ingesting data from the [IterableTables](https://github.com/queryverse/IterableTables.jl) interface would open a whole host of data sources from various Julia packages,

Separate the Soss compiler into its own package

1

Currently Soss bundles many dependencies in to make the full analysis convenient. Maybe it is possible to separate the compiler into a separate lightweight package that other packages can use...

Implement dummy types for distributions

2

Distributions.jl is not very flexible with respect to argument types. Implementing custom types, which are basically just wrappers for the arguments would allow more flexibility. For example, the following abuse...

Riemannian Manifold HMC

1

You [mentioned](https://discourse.julialang.org/t/mcmc-landscape/25654/40) having a toy implementation for RHMC. Is the code available somewhere? I would like to experiment with it using [higher order AD for matrix expressions](http://www.matrixcalculus.org/matrixcalculus.pdf).

Omitting results to points themselves

8

I would like to get the `k` nearest points that are not the querying points when using the same data for building the tree and querying it. Is it possible...