subspace | Lightweight vector and matrix math library | Math library

In machine translation, we always need to slice out the first timestep (the SOS token) in the annotation and prediction.

When using batch_first=False, slicing out the first timestep still keeps the tensor contiguous.

I have binary classified data that exists in high dimensional space, but I have a large amount of options for which exact subspace I use to represent the data. How do I calculate how well binary data clusters in a given high dimensional space?

Here is an example of one particular high dimensional representation reduced to 2D with t-sne (for visualization purposes, the data persists in high D space)

I would like a functional way to say "this space was x good at seperating True and False data", so that this spacial representation of the data can be compared to some other spacial representation.

https://docs.ipfs.io/guides/concepts/dnslink/

I think ipns uses a hash of your public peer id or something to create unique namespaces. I see no comment about how this works for DNSLink. What are the subspaces for DNSLink? If there is a global namespace as the documents suggest, how is this handled?

Currently, I have a tensor A, and a tensor U where U is an orthogonal matrix and is of full rank(so that its columns is a set of basis of U's column space, and all columns, say, u_i, have a norm of 1).

I am trying to compute the projection of each row of A onto the column space of U, using the formula from this post.

Which is, to compute Proj(A). Is there any convenient functions or better operations to achieve this? Thanks.

Say I have a batch of images in the form of tensors with dimensions (B x C x W x H) where B is the batch size, C is the number of channels in the image, and W and H are the width and height of the image respectively. I'm looking to use the transforms.Normalize() function to normalize my images with respect to the mean and standard deviation of the dataset across the C image channels, meaning that I want a resulting tensor in the form 1 x C. Is there a straightforward way to do this?

I tried torch.view(C, -1).mean(1) and torch.view(C, -1).std(1) but I get the error:

for a specific data analysis, I use the pre-compiled ica_linux binary from this repository. In the README.md of that repository, the authors state that you can achieve a major speedup when compiling the binary with architecture specific BLAS/LAPACK implementations. In our case, the speedup might actually save days of computing time, so i am very eager to compile the file with the appropriate libraries. Now, I don't have a lot of experience with makefiles, so I am a bit confused by the process -- which is why I am asking for your help.

My system is a HP ProLiant d360 Server with two Intel Xeon E5-2650 v3 deca-core processors and Ubuntu 16.04.6 x64

I have some question how to speed up my code.

I have a shp with more than 100000 polygons and I want calculate minimal bounding box for each of them. I found some code for this part here: https://rdrr.io/cran/flightplanning/src/R/utils.R

Code calculate min bounding box from coordinates in matrix form for one poylgon at time.

I use loop and data.table but for bigger shp (50000+ polygons) is this process really time consuming.

Is maybe possible use getMinBBox() with lapply inside data.table?

thank you

my code:

I have a multiple dimensions tensor, let's take this simple one as example:

I have an strange issue with a WPF TreeView in C#. When running my App i want to expand and collapse all nodes from the selected TreeView node. When i select a root node, the IsSelected property is not updated. When i manually expand the root node and click on the child node, the IsSelected property is updated. The problem is that when i use the context menu to expand and collapse the child nodes on the root node, the root node is not expanded. When i expand the root node after using the context menu i can see that all child nodes have been expanded correctly.

I have searched a lot in the internet, but couldn't find a similar issue. Furthermore i read a lot of tutorials and howtos on C# WPF TreeViews. So far my implementation seems to me to be correct, thats why i don't understand why the root property IsSelected is not set when clicking on the node.

Can anyone help me? I have implemented the TreeView, ViewModel and Model as follows. Thanks in advance

BR

Michael