tommelt

Hi, I am using the function `basis_function_ders_one` to generate B-spline basis functions for my project. I notice that they are defined between the knots ti

can we add support for build systems

1

@isazi we spoke at the RSECon23 workshop. I have a simple example fortran code ([Poisson-Solver](https://github.com/TomMelt/Poisson-Solver)) which uses: * `openMP` GPU offloading * `openACC` GPU offloading * as well as their...

feature request - Energy profiling

14

Hi, ArmForge has a feature ([perf-report](https://developer.arm.com/documentation/101136/2103/Performance-Reports/Interpret-performance-reports/Energy-breakdown?lang=en)) that can estimate power usage of a binary.  Is it possible to do something like this in omnitrace? I had tried using AMDuProf...

possible unexpected behaviour for `impl_thrust_rtc/storage.py`

2

https://github.com/open-atmos/PySDM/blob/64a54694dad9ee4d817498ec1fc845716ea184cc/PySDM/backends/impl_thrust_rtc/storage.py#L139-L162 I think `impl_thrust_rtc/storage.py` is possibly not behaving as expected. For example: truedivision in python will give `int(3)/int(2) == int(1.5)` But I think cuda code uses c/c++ which gives `int(3)/int(2)...

__isub__ implementation is missing scalar option

3

https://github.com/open-atmos/PySDM/blob/64a54694dad9ee4d817498ec1fc845716ea184cc/PySDM/backends/impl_numba/storage.py#L60-L69 `__isub__` is missing scalar implementation (compare with `__iadd__` above)

test code fails on first run

1

**Bug summary** When I run [this code](https://github.com/codeplaysoftware/syclacademy/blob/1b8f1d1778b00d1f319af90f87502a9f18d9b1ec/Code_Exercises/Introduction_to_SYCL/solution.cpp) from the codeplay demo, it fails on first run. The second time it runs fine. **To Reproduce** Steps to reproduce the behavior. _If...

should we add static analysis to our repo?

4

Should we use something like [sonarqube](https://www.sonarsource.com/products/sonarqube/) to do static analysis on the fortran source code?

overload `torch_tensor_delete`

1

overload `torch_tensor_delete` so we no longer have to delete each element of input/output tensor manually. It could just take the array and loop over the bounds :ok_hand: https://github.com/Cambridge-ICCS/FTorch/blob/490272ef5662936be26697d496b14f6a281214fa/src/ftorch.f90#L274-L286

Add intel ifx and ifort build to CI

1

Probably make a docker image with intel compilers already installed

Melt windows fix

1

fixes #124 This is currently wip. It needs to be tested by end user. If this fixes the problem I will tidy up the int64 issue and we can merge...