projectchrono/dem-engine

A dual-GPU DEM solver with complex grain geometry support

dem-engine's Issues

can't run the demo examples

Description

I am compiling DEME with CUDA 12.0 and Visual Studio 2019, and I am trying to run the demo examples in debug mode. Initially, I encountered a runtime error:

The variable 'new_stream' is being used without being initialized

I uncommented the following line where the error occurred:

cudaStreamCreate(&new_stream);

When I ran the program again, I encountered another error:

Unhandled exception at 0x00007FFEF9A632CF (nvcuda64.dll) (in DEMdemo_BallDrop2D.exe): 0xC0000006: An error in the page (status code 0x0000000900960118) occurred while reading at position 0x0000000900960118.

Steps to Reproduce

1. Compile DEME with CUDA 12.0 and Visual Studio 2019.
2. Run the demo examples in debug mode.
3. Uncomment the line cudaStreamCreate(&new_stream);.
4. Run the program again.

Expected Behavior

The demo should run without causing runtime errors or exceptions.

Actual Behavior

An unhandled exception occurs at runtime, specifically in nvcuda64.dll.

Error Messages

• Initial error:

  The variable 'new_stream' is being used without being initialized
  

• After uncommenting cudaStreamCreate(&new_stream);:

  Unhandled exception at 0x00007FFEF9A632CF (nvcuda64.dll) (in DEMdemo_BallDrop2D.exe): 0xC0000006: An error in the page (status code 0x0000000900960118) occurred while reading at position 0x0000000900960118.
  

Environment

• CUDA Version: 12.0
• Visual Studio Version: 2019
• Operating System: [Windows]
• GPU: [GTX 1650 Ti]

Additional Information

I am unsure if this error is due to incorrect code or insufficient device computing power.

a problem for installation

when i install the python package of DEME, there is an error that "coult not build wheels for deme, which is required to install pyproject.toml-based projects". could any one give me a help? thanks!

PDSampler should be random

The DEM-Engine implementation of PDSampler seems to have a flaw: every time we call it again, it has the same random state.
Indeed, when the user calls the SampleCylinderZ for instance (observed with other sampling functions), the same set of random positions is sampled. The minimal example shows the issue.

import DEME
import matplotlib.pyplot as plt
for i in range(10):
    sampler = DEME.PDSampler(0.1)
    pos = sampler.SampleCylinderZ([0,0,0], 0.5, 0)
    x = [p[0] for p in pos]
    y = [p[1] for p in pos]
    plt.scatter(x,y, label=i, alpha=0.2)
    print(len(pos))
plt.legend(loc='center left', bbox_to_anchor=[1.05,0.5], title='Batch #')

The printed text is a series of "56", meaning that the length of the sampled positions is always the same.
The obtained figure is the following:

We can clearly see that each set of points overlaps the others. The expected behavior would be something like the one shown in a (less efficient) Python sampler I made (prints 55, 58, 53, 54, ... as lengths):

Additional suggestion: an additional boolean argument to DEME.PDSampler called (or similar to) keep_history that would allow the users to choose if they want to keep the previously sampled positions or erase the history every time a new sampling is applied.

Please add Dockerfile to repository

Hi, thank you for this project!

Since you mention the Docker container is available, could you please add the source Dockefile?