Auto_ML_C 0.0.10
Illustrate:
这是崔连山和小伙伴们的机器学习拓展包,代有浓厚的社会主义开源分享精神,极富创造力和战斗力。在这里让我们为他们鼓掌 🥂
Spend
测试集数据位置:{Example}
配置如下:
| Windows | Windows | MacOS | Linux | |
|---|---|---|---|---|
| 型号 | i7-9750H | i7-9750H | M1 | E5-2640 V4 |
| 核心 | 6核心12线程 | 6核心12线程 | 8核心 | 20核心40线程 |
| 频率 | 2.67GHz | 3.2GHz | 2.40GHz |
运行速度对比结果如下:
| 未集成 | Windows1 | Windows2 | MacOS | Linux |
|---|---|---|---|---|
| ALL_FUNCTION | 47.364 | 43.681 | 34.013 | ==27.282== |
| binary_ROC() | 45.809 | 42.964 | 32.751 | ==32.143== |
| auto_model() | 53.498 | 48.649 | ==38.267== | 40.794 |
| estimator_violion() | 1.191 | ==1.021== | 1.678 | 2.395 |
| 集成 | Windows1 | Windows2 | MacOS | Linux |
|---|---|---|---|---|
| binary_ROC() | 46.2 s | 43.2 s | CPU times: user 4.82 s, sys: 365 ms, total: 5.18 s Wall time: ==32.9s== | CPU times: user 9.59 s, sys: 3.89 s, total: 13.5 s Wall time: 33.3 s |
| auto_model() | 50.4 s | 47.1 s | CPU times: user 9.75 s, sys: 247 ms, total: 10 s Wall time: ==38.1 s== | CPU times: user 15.1 s, sys: 1.68 s, total: 16.8 s Wall time: 41.1 s |
| estimator_violion() | 1.16 s | Wall time: ==1.01 s== | CPU times: user 2.02 s, sys: 70.1 ms, total: 2.09 s Wall time: 1.69 s | CPU times: user 3.85 s, sys: 2.32 s, total: 6.17 s Wall time: 2.23 s |
Request_install
可以参考学习当前目录下的环境备份:Auto_ML_C.yaml
主要是涉及到的软件如下:
| Package | 最低版本——待检测 |
|---|---|
| python=3.8.10 | |
| seaborn=0.11.2 | |
| pandas=1.3.3 | |
| matplotlib=3.4.2 | |
| numpy=1.20.3 |
Content:
该包是基于Sklearn,imblance等机器学习拓展包之上的Package,共计划分为两个部分,
-
分类任务
-
binary_classfication.py
内部可用函数如下
函数名 功能 返回值 cal_add_1(num1,num2)👋 简单的欢迎函数 num1,num2 LogisticRegressionCV_mdoel(X, Y,cv) SGDClassifier_model(X,Y,cv) LinearDiscriminantAnalysis_model(X, Y,cv) LinearSVC_model(X, Y,cv) SVC_model(X, Y,cv) DecisionTreeClassifier_model(X,Y,cv) AdaBoostClassifier_model(X,Y,cv) BaggingClassifier_model(X, Y,cv) GradientBoostingClassifier_model(X, Y,cv) RandomForestClassifier_model(X, Y,cv) KNeighborsClassifier_model(X, Y,cv) BernoulliNB_model(X, Y,cv) GaussianNB_model(X,Y,cv) 下面是总函数 binary_ROC(X,Y,k,fig_name) 绘制标量超参数搜索下最佳的ROC fig auto_model(X, Y, k) 模型的标量超参数搜索结果 Auc_data, Acc_data,
Recall_data, Precision_dataestimator_violion(df1,df2,fig_name) 为auto_model结果的Dataframe绘制小提琴图 fig -
多分类函数
等待
-
特征筛选函数Feature_struction
-
waited
-
How to Use
Installation
# Method 1
# Create a new environment, here is conda as an example
conda create --name Auto_ML_C python=3.8.10
# Activate the newly created environment
conda activate Auto_ML_C
# Installation package
pip install Auto_ML_C==0.0.8
# Suggest the pipeline of Jupyter notebook [optional, recommended]
conda install jupyter notebook
conda install ipykernel
python -m ipykernel install --user --name Auto_ML_C --display-name "Auto_ML_C"
# Install Sklearn 0.6. this will fixed next version
conda install -c conda-forge sklearn-contrib-lightning
# Method2
# Use the yaml environment file on the GitHub homepage to directly copy the current environment
conda env create -n Auto_ML_C -f Auto_ML_C.yaml
# Activate the newly created environment
conda activate
# Suggest the pipeline of Jupyter notebook [optional, recommended]
conda install jupyter notebook
conda install ipykernel
python -m ipykernel install --user --name Auto_ML_C --display-name "Auto_ML_C"
Feature_struction
#
Binary Classication
# Here is an example of the function binary_classfication_ws
# 这里以函数binary_classfication_ws举例
# 开始加载环境
import pandas as pd
import numpy as np
import auto_ml_c.binary_classfication as abc
# 读取测试数据
df = pd.read_csv("2_data_deal_smote.csv")
X = df.iloc[:,:-1]
Y = df["label"]
score = 'accuracy'
# The first function, draw ROC image
tmp_a = abc.binary_ROC(X,Y,cv,"111","accuracy")
# The second function, get Auc_data, Acc_data, Recall_data, Precision_data
tmp_b1,tmp_b2,tmp_b3,tmp_b4 = abc.auto_model(X,Y,cv,"accuracy")
# The third function, draw the evaluation graph obtained by function 2 auto_model
tmp_c = abc.estimator_violion(tmp_b1,tmp_b2,"Violionplot")
ConTact
VX:Cuizy13390906310_ic
QQ:1776228595
E-mail:1776228595@qq.com
GitHub:地址待填写
Metadata
Release files for Auto-ML-C 0.0.11
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| Auto_ML_C-0.0.11.tar.gz | 12.5 kB | Without provenance | Details for Auto_ML_C-0.0.11.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| Auto_ML_C-0.0.11-py3-none-any.whl | Python 3 | none | any | Details for Auto_ML_C-0.0.11-py3-none-any.whl |
Total release size: 25.5 kB
Release files / Auto_ML_C-0.0.11.tar.gz
| Download URL | Auto_ML_C-0.0.11.tar.gz |
|---|---|
| Size | 12.5 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
35148ef84e7db4d3f58f0cd7aef8e66a00d43f8f689fc62c6b620a8dacdf04cc
|
|
BLAKE2b-256 checksum How to use checksums |
34762994c8eee02e4aee030d19780ff548321965e5730546172025e36b87dc62
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/3.4.2 importlib_metadata/4.8.1 pkginfo/1.7.1 requests/2.26.0 requests-toolbelt/0.9.1 tqdm/4.62.3 CPython/3.8.0
|
Release files / Auto_ML_C-0.0.11-py3-none-any.whl
| Download URL | Auto_ML_C-0.0.11-py3-none-any.whl |
|---|---|
| Size | 13.0 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
50e0e33ff104608f32fc0d67610d4a9f6091110ee87e96d3c10c71916aa0e55d
|
|
BLAKE2b-256 checksum How to use checksums |
6b6643e12d0c40b15d53826f2f7c8422b6b3d4a973b42758b6fb8409666bddb4
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/3.4.2 importlib_metadata/4.8.1 pkginfo/1.7.1 requests/2.26.0 requests-toolbelt/0.9.1 tqdm/4.62.3 CPython/3.8.0
|
Security for Auto-ML-C 0.0.11
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.