معرفی شرکت ها
PyBenchFCN-1.0.3
تبلیغات ما
مشتریان به طور فزاینده ای آنلاین هستند. تبلیغات می تواند به آنها کمک کند تا کسب و کار شما را پیدا کنند.
مشاهده بیشتر
تبلیغات ما
مشتریان به طور فزاینده ای آنلاین هستند. تبلیغات می تواند به آنها کمک کند تا کسب و کار شما را پیدا کنند.
مشاهده بیشتر
تبلیغات ما
مشتریان به طور فزاینده ای آنلاین هستند. تبلیغات می تواند به آنها کمک کند تا کسب و کار شما را پیدا کنند.
مشاهده بیشتر
تبلیغات ما
مشتریان به طور فزاینده ای آنلاین هستند. تبلیغات می تواند به آنها کمک کند تا کسب و کار شما را پیدا کنند.
مشاهده بیشتر
تبلیغات ما
مشتریان به طور فزاینده ای آنلاین هستند. تبلیغات می تواند به آنها کمک کند تا کسب و کار شما را پیدا کنند.
مشاهده بیشترتوضیحات
A python implementation of optimization benchmarks
| ویژگی | مقدار |
|---|---|
| سیستم عامل | OS Independent |
| نام فایل | PyBenchFCN-1.0.3 |
| نام | PyBenchFCN |
| نسخه کتابخانه | 1.0.3 |
| نگهدارنده | [] |
| ایمیل نگهدارنده | [] |
| نویسنده | Yifan He |
| ایمیل نویسنده | he.yifan.xs@alumni.tsukuba.ac.jp |
| آدرس صفحه اصلی | https://github.com/Y1fanHE/PyBenchFCN |
| آدرس اینترنتی | https://pypi.org/project/PyBenchFCN/ |
| مجوز | - |
<h1>
<p align="center"># PyBenchFCN #</p>
<p align="center">A python implementation of optimization benchmarks</p>
</h1>
<p align="center">
<img src="./image/f51_3D.png" width=300><img src="./image/f51_2D.png" width=300></p>
<p align="center"><img src="./image/f58_2D.png" width=300><img src="./image/f58_3D.png" width=300></p>
- [How to Install](#how-to-install)
- [How to Use](#how-to-use)
- [Set Benchmark Function](#set-benchmark-function)
- [Plot Fitness Landscape](#plot-fitness-landscape)
- [List of Functions](#list-of-functions)
- [Classical Single-Objective Optimization](#classical-single-objective-optimization)
- [Discrete Optimization](#discrete-optimization)
- [Multi-Objective Optimization](#multi-objective-optimization)
- [Real-World Optimization](#real-world-optimization)
- [Authors](#authors)
- [License](#license)
- [Acknowledgement](#acknowledgement)
## How to Install
This library is a python implementation for the MatLab package [BenchmarkFcns Toolbox](http://benchmarkfcns.xyz/).
You can simply install with command ```pip install PyBenchFCN```.
- Pre-request: ```numpy```, ```matplotlib```
## How to Use
The input of each numerical optimization problem could be a 1-D ndarray, or 2-D ndarray.
- **1-D array**
- an example of **a solution (individual)** for 10D problem is ```np.random.uniform(0, 1, 10)```, where each entry is a decision variable.
- <u>use ```f()``` to return a fitness value (scalar for SOP, 1D-array for MOP).</u>
- **2-D array**
- an example of **group of solutions (population)** for 10D problem is ```np.random.uniform(0, 1, (5, 10))```, where each row (totally 5) is an individual.
- <u>use ```F()``` to return an array of fitness value (1-D array for SOP, 2-D array for MOP).</u>
### Set Benchmark Function
To set a benchmark function, one may see the sample code in ```Factory.py``` in the [repository](https://github.com/Y1fanHE/PyBenchFCN), or follow the script below. Also, there is a sample optimization program provided in ```sample.py```.
```python3
import numpy as np
# import single objective problems from PyBenchFCN
from PyBenchFCN import SingleObjectiveProblem as SOP
n_var = 10 # dimension of problem
n_pop = 3 # size of population
problem = SOP.ackleyfcn(n_var) # Ackley problem
'''same function as the code above
from PyBenchFCN import Factory
problem = Factory.set_sop("f1", n_var)
'''
print( np.round(problem.optimalF, 5) ) # show rounded optimal value
xl, xu = problem.boundaries # bound of problem
x = np.random.uniform(xl, xu, n_var) # initialize a solution
print( problem.f(x) ) # show fitness value as scalar
X = np.random.uniform( xl, xu, (n_pop, n_var) ) # initialize a population
print( problem.F(X) ) # show fitness values as 1d-array
```
### Plot Fitness Landscape
To plot a fitness landscape (2D space), one can use the code below. **Notice, this function only works for continuous SOPs.**
```python3
from PyBenchFCN import Tool
Tool.plot_sop("sphere", mode="save") # plot and save landscape of Sphere function
Tool.plot_sop("schwefel", plot_type="contour") # plot contour plot of Schwefel function
```
## List of Functions
### Classical Single-Objective Optimization
Totally, [61 single-objective functions](./SingleObjectiveProblem.md) are implemented. The plot of 2D versions of 59 problems are provided. Please check the homepage of [BenchmarkFcns Toolbox](http://benchmarkfcns.xyz/) or [this website](https://www.sfu.ca/~ssurjano/optimization.html) for the further documentation.
### Discrete Optimization
*Under development ...*
### Multi-Objective Optimization
*Under development ...*
### Real-World Optimization
*Under development ...*
## Authors
[Yifan He](https://y1fanhe.github.io) @ Dept. of CS, UTsukuba
## License
This project is licensed under the MIT License - see the [LICENSE](./LICENSE) file for details.
## Acknowledgement
PyBenchFCN is maintained by [Yifan He](https://y1fanhe.github.io). The author of this repostory is very grateful to Mr. Mazhar Ansari Ardeh, who implemented the MatLab package BenchFCNs Toolbox.
- If you find any mistakes, please report at a new issue.
- If you want to help me implement more benchmarks (discrete optimization, multi-objective optimization), please contact at [he.yifan.xs@alumni.tsukuba.ac.jp](mailto:he.yifan.xs@alumni.tsukuba.ac.jp).
نیازمندی
| مقدار | نام |
|---|---|
| - | numpy |
| - | matplotlib |
زبان مورد نیاز
| مقدار | نام |
|---|---|
| >=3.6 | Python |
نحوه نصب
نصب پکیج whl PyBenchFCN-1.0.3:
pip install PyBenchFCN-1.0.3.whl
نصب پکیج tar.gz PyBenchFCN-1.0.3:
pip install PyBenchFCN-1.0.3.tar.gz