Igor Krawczuk

Thanks:-)

I want to study the effect of modifying edge weights and adding/removing edges across different GNNs, for this I need the gradient w.r.t to edges. (i.e., I want to check...

thank your very much for the information. But to clarify, if I use a weighted graph with e.g. $A=[0 1 0.5; 1 0 0 ;0.5 0 0]$, I will be...

D'accord, is there a place where I can find out which parts are fully obsolete and which ones are still useful (e.g. the explicit signal_connect?) Then I'd try my hand...

Small update: I have changed the CMakeLists.txt entry to ```cmake find_package(QuickQanava REQUIRED) target_link_libraries(${PROJECT_NAME} PRIVATE QuickQanava ) ``` and have verified that `make install` placed the qmldir in the correct folder....

Hi, this for example is the comment link: https://www.reddit.com/r/programming/comments/7z59kg/opensource_project_which_found_12_bugs_in/ for the submission link http://ithare.com/c17-compiler-bug-hunt-very-first-results-12-bugs-reported-3-already-fixed/

mainly asking to clarify (and for the publics benefit I guess): [I see torch 2.0](https://pytorch.org/docs/stable/tensors.html) has `scatter`,`scatter_add` etc. , but for the `scatter_max`and other reductions, for an in-place replacement one...