SujayP

Results 10 comments of SujayP

Hey, It's how the "Leaky, integrate and fire" neuron model is described. Neuron potential decays every timestep. You can refer to the following book to get a deeper understanding: "https://neuronaldynamics.epfl.ch/online/Ch5.S2.html"

Yes, I can take that up. Thanks! Can you start a branch for me?

Got it! Thanks @mathurinm

Hey @kmedved and @mathurinm, I did get started on it but need to finish some work for my paper submission in April. I will start working on this as soon...

Hi @mathurinm and @kmedved Apologies for the delay in communication. I am starting afresh now. Before I make changes, I wanted to get a confirmation of my plan. For now,...

Hi @QB3 Thanks for the prompt feedback! > * 1 First, add the `sample_weights` to the `Quadratic` class, note that you will also have to change the Lipschitz constant computation...

@mathurinm and @QB3, In this case, we can go for separate classes. I wasn't aware of this numba issue. I have a question, The changes in #245 and this issue...

@mathurinm I will push my changes by Thursday EOD. If I am not able to I will hand it over to you guys. Apologies for the delay.

Hi @mathurinm @kmedved @QB3 I have added a new class 'WeightedQuadratic' in datafits/single_task.py and created a PR. I have tested it with examples/plot_sparse_recovery.py and it seems to work fine (with...