跳到主要内容

孤立森林模板

https://scikit-learn.org/stable/auto_examples/ensemble/plot_isolation_forest.html#sphx-glr-auto-examples-ensemble-plot-isolation-forest-py

下图为项目概览： image.png 项目分为四部分：

  • 第一部分：数据生成，生成用于训练以及预测的数据
import numpy as np
import pandas as pd

import suanpan
from suanpan.app import app
from suanpan.app.arguments import Csv


def toDataFrame(X):
    featureColumns = ["feature_{}".format(i) for i in range(X.shape[1])]
    return pd.DataFrame(X, columns=featureColumns)


@app.output(Csv(key="outputData1"))
@app.output(Csv(key="outputData2"))
@app.output(Csv(key="outputData3"))
def Demo(context):
    args = context.args

    rng = np.random.RandomState(42)

    # Generate train data
    X = 0.3 * rng.randn(100, 2)
    X_train = np.r_[X + 2, X - 2]
    # Generate some regular novel observations
    X = 0.3 * rng.randn(20, 2)
    X_test = np.r_[X + 2, X - 2]
    # Generate some abnormal novel observations
    X_outliers = rng.uniform(low=-4, high=4, size=(20, 2))

    return toDataFrame(X_train), toDataFrame(X_test), toDataFrame(X_outliers)


if __name__ == "__main__":
    suanpan.run(app)

  • 第二部分：孤立森林节点，接收第一部分生成训练数据，训练模型
    • 模型参数：max_samples=100, contamination=0.1
  • 第三部分：Decision Function输出，使用模型的decision_function功能，输出结果
import joblib
import numpy as np

import suanpan
from suanpan.app import app
from suanpan.app.arguments import File, Npy


@app.input(
    File(
        key="inputModel1", alias="inputModel", name="model", type="model", required=True
    )
)
@app.output(Npy(key="outputData1"))
@app.output(Npy(key="outputData2"))
@app.output(Npy(key="outputData3"))
def Demo(context):
    args = context.args

    model = joblib.load(args.inputModel)

    xx, yy = np.meshgrid(np.linspace(-5, 5, 50), np.linspace(-5, 5, 50))
    Z = model.decision_function(np.c_[xx.ravel(), yy.ravel()])
    Z = Z.reshape(xx.shape)

    return xx, yy, Z


if __name__ == "__main__":
    suanpan.run(app)

  • 第四部分：结果展示
import os

from matplotlib import pyplot as plt

import suanpan
from suanpan.app import app
from suanpan.app.arguments import Csv, Folder, Npy

TMP_FOLDER = "/tmp/result"


@app.input(Csv(key="inputData1", alias="X_train"))
@app.input(Csv(key="inputData2", alias="X_test"))
@app.input(Csv(key="inputData3", alias="X_outliers"))
@app.input(Npy(key="inputData4", alias="xx"))
@app.input(Npy(key="inputData5", alias="yy"))
@app.input(Npy(key="inputData6", alias="Z"))
@app.output(Folder(key="outputData1"))
def Demo(context):
    args = context.args

    X_train = args.X_train.values
    X_test = args.X_test.values
    X_outliers = args.X_outliers.values

    if not os.path.exists(TMP_FOLDER):
        os.makedirs(TMP_FOLDER)

    plt.title("IsolationForest")
    plt.contourf(
        args.xx, args.yy, args.Z, cmap=plt.cm.Blues_r  # pylint: disable=no-member
    )

    b1 = plt.scatter(X_train[:, 0], X_train[:, 1], c="white", s=20, edgecolor="k")
    b2 = plt.scatter(X_test[:, 0], X_test[:, 1], c="green", s=20, edgecolor="k")
    c = plt.scatter(X_outliers[:, 0], X_outliers[:, 1], c="red", s=20, edgecolor="k")
    plt.axis("tight")
    plt.xlim((-5, 5))
    plt.ylim((-5, 5))
    plt.legend(
        [b1, b2, c],
        [
            "training observations",
            "new regular observations",
            "new abnormal observations",
        ],
        loc="upper left",
    )
    plt.savefig(os.path.join(TMP_FOLDER, "isolation_forest.png"), format="png")

    return TMP_FOLDER


if __name__ == "__main__":
    suanpan.run(app)

效果如下：isolation_forest.png