vboussange

Assortativity extended to graphs with attributes

5

Hi there, I propose an extension of method `assortativity` to account for scalar properties of vertices. This is also done in [networkX (python)](https://networkx.org/documentation/stable/reference/algorithms/generated/networkx.algorithms.assortativity.attribute_assortativity_coefficient.html#networkx.algorithms.assortativity.attribute_assortativity_coefficient). """ assortativity(g,labels) Similar to `assortativity(g)` except that...

New rich club metric

1

Hi there, I suggest to add the rich-club metric (cf [Wikipedia](https://en.wikipedia.org/wiki/Rich-club_coefficient) ) to beautiful LightGraphs.jl, under the community metrics. `rich_club(g,k)` Return the non-normalised [rich-club coefficient](https://en.wikipedia.org/wiki/Rich-club_coefficient) of graph `g`, with degree...

New s_metric method

1

Hi there, I propose to add a new metric, so called s-metric. This is also done in [networkX](https://networkx.org/documentation/stable/reference/algorithms/generated/networkx.algorithms.smetric.s_metric.html#networkx.algorithms.smetric.s_metric). `s_metric(g; norm=true)` Return the normalised s-metric of `g`. The s-metric is defined...

Problem with SplitODEProblem in more than one dimension

1

Hi, It seems that there is an issue solving a `SplitODEProblem` with initial conditions of type `u0::Array{Float64,2}`. I get the error : `BoundsError: attempt to access 200-element view(::Array{Float64,2}, :, 1)...

Allow user to change the channel axis for BatchNorm function and the likes

2

Proposition for issue #1664 This PR allows the user to customize the channel axis for normalisation functions (`BatchNorm`,` GroupNorm` and `InstanceNorm`). ## Example ```Julia using Flux channel_size = 3 channel_axis...

Integrate `NNStopping` and `NNKolmogorov ` in HighDimPDE.jl

3

`NeuralPDE.KolmogorovPDEProblem`, documented in - https://neuralpde.sciml.ai/dev/solvers/kolmogorovbackwards_solver/ - https://neuralpde.sciml.ai/dev/examples/kolmogorovbackwards/ are `HighDimPDE.PIDEProblem` where the nonlinear term is 0. Associated code and documentation should be migrated to HighDimPDE.jl, which is a more appropriate package...

https://github.com/SciML/NeuralPDE.jl/issues/536

For now gradients `∇v_y,∇v_z` cannot be used in the nonlinear function ``` f(y,z,v_y,v_z,∇v_y,∇v_z, p, t) ``` Uncommenting those two line https://github.com/vboussange/HighDimPDE.jl/blob/21a25b7f13f15332a1b5872961315b7bcedc5a5b/src/DeepSplitting.jl#L92-L93 should make this work. Nonetheless, possibly because of issue...

Implementation of a `merge` function

4

Hey there, I thought it'd be very cool to have a specialisation of the `merge` function for `ComponentArray`s, with a similar behaviour as with `Dictionary`s or `NamedTuple`s. For instance, consider...

When using e.g. `ADAM` optimiser, one can use a `NCycle` iterator, i.e. ```julia maxiters_wanted = 1000 train_loader = Flux.Data.DataLoader((0:100,); batchsize = 4, shuffle = true) data = ncycle(train_loader, maxiters_wanted) res...