<!-- LTeX: language=en-US -->
# CuSIR
## Introduction
CuSIR is a Python code built on top of CuPy, a NumPy-like library for GPU-accelerated computing. It provides a solver for the two-dimensional diffusive SIR model, described by the following system of reaction-diffusion equations:
$$
\begin{align}
\partial_t S &= -\beta_{\mathbf{r}} S I - \gamma I + D_I \nabla^2 I \\\\\\\\
\partial_t I &= \beta_{\mathbf{r}} S I + D_S \nabla^2 S - \mathbf{v} \cdot \nabla I,
\end{align}
$$
where $S$ is the density of susceptible individuals, $I$ is the density of infected individuals, $\beta_{\mathbf{r}}$ is the transmission rate that depends on the location $\mathbf{r}$, $\gamma$ is the recovery/removal rate, $D_I$ and $D_S$ are diffusion coefficients, and $\mathbf{v}$ is the convection field.
## Requirements
To use the CuSIR package, you will need the following software and hardware:
- A CUDA-compatible GPU: A graphics processing unit (GPU) that supports CUDA. Check the list of CUDA-compatible GPUs on the NVIDIA website (https://developer.nvidia.com/cuda-gpus) to see if your GPU is supported.
- CUDA Toolkit: A parallel computing platform and programming model developed by NVIDIA for general-purpose computing on GPUs. You can download CUDA from the NVIDIA website (https://developer.nvidia.com/cuda-downloads).
*Note: Currently (January, 2023), the last version of CUDA (12) is not supported by CuPy. You will need to install any previous version of CUDA (recommended 11.2) to use CuSIR.*
- CuPy: A NumPy-like library for GPU-accelerated computing. You can install CuPy by following the instructions in the CuPy documentation (https://docs-cupy.chainer.org/en/stable/install.html).
## Installation
To install the CuSIR package, you can use pip by running the following command in your command prompt or terminal:
```bash
pip install cusir
```
This command will install the latest version of the CuSIR package. It is mandatory to meet the requirements listed above for CuSIR to work properly.
## Usage
Almost everything is implemented in the `system` class, which is located in the `system` module. The following code shows how to use the `system` class to solve the diffusive SIR model:
```python
import cusir.system as cs
# Define the spatial domain
Lx = 2**10
Ly = 2**10
# Create the system object
s = cs.system(Lx, Ly)
# Define the system parameters
s.beta = 1 # Transmission rate
s.gamma = 0.1 # Recovery/removal rate
s.D_I = 1 # Diffusion coefficient for infected individuals
s.D_S = 1 # Diffusion coefficient for susceptible individuals
# Define the initial conditions
s.set_plane_initial_conditions()
# Solve the system
for _ in range(10000):
s.update() # Update the system
s.rigid_x() # Apply rigid boundary conditions in the x-direction
#You can also use the following to do the same:
#s.solve(10000)
# Get the solution
S = s.S.get() # get() pulls the data from the GPU to the CPU as a NumPy array
I = s.I.get()
```
## License
CuSIR is licensed under the MIT license. See the `LICENSE` file for more details.
Raw data
{
"_id": null,
"home_page": "",
"name": "cusir",
"maintainer": "",
"docs_url": null,
"requires_python": ">=3.7",
"maintainer_email": "",
"keywords": "sir,epidemiology,CuPy,GPU,CUDA,diffusion,reaction-diffusion,parallel computing",
"author": "",
"author_email": "renzozs <renzozagarrasaez@gmail.com>",
"download_url": "https://files.pythonhosted.org/packages/24/33/db72509660f415ce45d497ad840696b9d2057662ce85bac953d06f83e2e0/cusir-1.0.1.tar.gz",
"platform": null,
"description": "<!-- LTeX: language=en-US -->\n# CuSIR\n\n## Introduction \nCuSIR is a Python code built on top of CuPy, a NumPy-like library for GPU-accelerated computing. It provides a solver for the two-dimensional diffusive SIR model, described by the following system of reaction-diffusion equations:\n\n$$\n\\begin{align} \n\\partial_t S &= -\\beta_{\\mathbf{r}} S I - \\gamma I + D_I \\nabla^2 I \\\\\\\\\\\\\\\\\n\\partial_t I &= \\beta_{\\mathbf{r}} S I + D_S \\nabla^2 S - \\mathbf{v} \\cdot \\nabla I,\n\\end{align}\n$$\n\nwhere $S$ is the density of susceptible individuals, $I$ is the density of infected individuals, $\\beta_{\\mathbf{r}}$ is the transmission rate that depends on the location $\\mathbf{r}$, $\\gamma$ is the recovery/removal rate, $D_I$ and $D_S$ are diffusion coefficients, and $\\mathbf{v}$ is the convection field. \n\n## Requirements\n\nTo use the CuSIR package, you will need the following software and hardware:\n\n- A CUDA-compatible GPU: A graphics processing unit (GPU) that supports CUDA. Check the list of CUDA-compatible GPUs on the NVIDIA website (https://developer.nvidia.com/cuda-gpus) to see if your GPU is supported.\n\n- CUDA Toolkit: A parallel computing platform and programming model developed by NVIDIA for general-purpose computing on GPUs. You can download CUDA from the NVIDIA website (https://developer.nvidia.com/cuda-downloads). \n\n*Note: Currently (January, 2023), the last version of CUDA (12) is not supported by CuPy. You will need to install any previous version of CUDA (recommended 11.2) to use CuSIR.* \n\n- CuPy: A NumPy-like library for GPU-accelerated computing. You can install CuPy by following the instructions in the CuPy documentation (https://docs-cupy.chainer.org/en/stable/install.html).\n\n## Installation\n\nTo install the CuSIR package, you can use pip by running the following command in your command prompt or terminal:\n\n```bash\npip install cusir\n```\n\nThis command will install the latest version of the CuSIR package. It is mandatory to meet the requirements listed above for CuSIR to work properly.\n\n## Usage\n\nAlmost everything is implemented in the `system` class, which is located in the `system` module. The following code shows how to use the `system` class to solve the diffusive SIR model:\n\n```python\nimport cusir.system as cs\n\n# Define the spatial domain\nLx = 2**10\nLy = 2**10\n\n# Create the system object\ns = cs.system(Lx, Ly)\n\n# Define the system parameters\ns.beta = 1 # Transmission rate\ns.gamma = 0.1 # Recovery/removal rate \ns.D_I = 1 # Diffusion coefficient for infected individuals\ns.D_S = 1 # Diffusion coefficient for susceptible individuals\n\n# Define the initial conditions\ns.set_plane_initial_conditions()\n\n# Solve the system\nfor _ in range(10000):\n s.update() # Update the system\n s.rigid_x() # Apply rigid boundary conditions in the x-direction\n\n#You can also use the following to do the same:\n#s.solve(10000)\n\n# Get the solution\nS = s.S.get() # get() pulls the data from the GPU to the CPU as a NumPy array\nI = s.I.get()\n```\n\n## License\n\nCuSIR is licensed under the MIT license. See the `LICENSE` file for more details.\n\n\n\n\n",
"bugtrack_url": null,
"license": "",
"summary": "A SIR diffusive model based in CuPy",
"version": "1.0.1",
"split_keywords": [
"sir",
"epidemiology",
"cupy",
"gpu",
"cuda",
"diffusion",
"reaction-diffusion",
"parallel computing"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "922e87a6543c961c5b6341bc0ba41e7fc2afadf082fa5830e6afc84582e6b15e",
"md5": "08e5f0f4b2b34e62f753fc688e80483d",
"sha256": "9f36c4d43a9edd17c4fa47945f52b3ccc241775109dae6267a170bf3981c3300"
},
"downloads": -1,
"filename": "cusir-1.0.1-py3-none-any.whl",
"has_sig": false,
"md5_digest": "08e5f0f4b2b34e62f753fc688e80483d",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.7",
"size": 7502,
"upload_time": "2023-01-20T15:15:08",
"upload_time_iso_8601": "2023-01-20T15:15:08.243809Z",
"url": "https://files.pythonhosted.org/packages/92/2e/87a6543c961c5b6341bc0ba41e7fc2afadf082fa5830e6afc84582e6b15e/cusir-1.0.1-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "2433db72509660f415ce45d497ad840696b9d2057662ce85bac953d06f83e2e0",
"md5": "763ec618cab520e41662d09d8fae2a4d",
"sha256": "fb82e99f85a0b285834e3496b83a76bbc2b8990298eafe90c3741b8a03f29168"
},
"downloads": -1,
"filename": "cusir-1.0.1.tar.gz",
"has_sig": false,
"md5_digest": "763ec618cab520e41662d09d8fae2a4d",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.7",
"size": 9569,
"upload_time": "2023-01-20T15:15:10",
"upload_time_iso_8601": "2023-01-20T15:15:10.599592Z",
"url": "https://files.pythonhosted.org/packages/24/33/db72509660f415ce45d497ad840696b9d2057662ce85bac953d06f83e2e0/cusir-1.0.1.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-01-20 15:15:10",
"github": false,
"gitlab": false,
"bitbucket": false,
"lcname": "cusir"
}
JSON