https://github.com/ibbd-dev/python-ibbd-algo

Python算法模块
https://github.com/ibbd-dev/python-ibbd-algo

Last synced: 6 days ago
JSON representation

Python算法模块

• Host: GitHub
• URL: https://github.com/ibbd-dev/python-ibbd-algo
• Owner: ibbd-dev
• License: apache-2.0
• Created: 2020-03-26T06:54:50.000Z (over 4 years ago)
• Default Branch: master
• Last Pushed: 2022-02-26T07:53:18.000Z (over 2 years ago)
• Last Synced: 2023-03-06T13:05:57.812Z (over 1 year ago)
• Language: Python
• Size: 239 KB
• Stars: 1
• Watchers: 2
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# python-ibbd-algo

IBBD常用算法

## Install

```sh

# 安装依赖

pip3 install -r https://github.com/ibbd-dev/python-ibbd-algo/raw/master/requirements.txt

# 安装

pip3 install git+https://github.com/ibbd-dev/python-ibbd-algo.git

```

## 支持的算法列表

### 序列相关函数

从两个序列中找到得分最高的匹配

```python

from ibbd_algo.sequence import Match

```

### Optics

scikit-learn中包含很多聚类的算法，但是在使用的过程中发现一个比较大的问题，如optics算法，不能自定义距离，只能造一个轮子。

```python

from ibbd_algo.optics import Optics

points = [

    np.array((1, 1)),

    np.array((1, 3)),

    np.array((2, 2)),

    np.array((4, 6)),

    np.array((5, 7)),

]

# 默认使用欧氏距离

optics = Optics(4, 2)

optics.fit(points)

labels = optics.cluster(2)

print(labels)

# 输出：[0 0 0 1 1]

```

可以使用自定义距离：

```python

def distance(point1, point2):

    data = [abs(a-b) for a, b in zip(point1, point2)]

    return sum(data)

optics = Optics(4, 2, distance=distance)

optics.fit(points)

```