Python And sklearn2pmml：sklearn2pmml Introduction to library functions 、 install 、 A detailed introduction to how to use

Catalog

sklearn2pmml Introduction to library functions

1、 A typical workflow summary

sklearn2pmml Installation of library functions

sklearn2pmml How to use library functions

1、 A simple decision tree model for Iris species classification

2、 A more refined logistic regression model

sklearn2pmml Introduction to library functions

        sklearn2pmml It's used to put Scikit The learning pipeline is transformed into PMML Of Python library . This library is JPMML-SkLearn A thin wrapper for command line applications . List of supported evaluators and converter types , Please refer to JPMML-SkLearn characteristic .

1、 A typical workflow summary

• Create a PMMLPipeline object , And fill it with piping steps as usual . class sklearn2pmml.pipeline.PMMLPipeline Expanded sklearn.pipeline class . Pipes have the following functions :
• If PMMLPipeline.fit(X, y) The method is to use panda Called .DataFrame Or panda .Series Object as X Parameters , Then its column name is used as the property name . otherwise , The default feature name is “x1”、“x2”,..“x {number_of_features}”.
• If PMMLPipeline.fit(X, y) The method is to use panda Called .Series Object as y Parameters , Then use its name as the target name ( For the supervision model ). otherwise , The target name defaults to “y”.
• Install and verify as usual pipeline.
• Optionally , By calling PMMLPipeline.verify(X) Method to calculate validation data and embed it into PMMLPipeline In the object , This method uses a small but representative subset of training data .
• By calling utility methods sklearn2pmml, take PMMLPipeline Object to the local file system PMML file .pmml_destination_path sklearn2pmml(pipeline).

sklearn2pmml Installation of library functions

pip install sklearn2pmml
pip install --user -i https://pypi.tuna.tsinghua.edu.cn/simple sklearn2pmml

sklearn2pmml How to use library functions

1、 A simple decision tree model for Iris species classification

import pandas
iris_df = pandas.read_csv("Iris.csv")
iris_X = iris_df[iris_df.columns.difference(["Species"])]
iris_y = iris_df["Species"]
from sklearn.tree import DecisionTreeClassifier
from sklearn2pmml.pipeline import PMMLPipeline
pipeline = PMMLPipeline([
("classifier", DecisionTreeClassifier())
])
pipeline.fit(iris_X, iris_y)
from sklearn2pmml import sklearn2pmml
sklearn2pmml(pipeline, "DecisionTreeIris.pmml", with_repr = True)

2、 A more refined logistic regression model

import pandas
iris_df = pandas.read_csv("Iris.csv")
iris_X = iris_df[iris_df.columns.difference(["Species"])]
iris_y = iris_df["Species"]
from sklearn_pandas import DataFrameMapper
from sklearn.decomposition import PCA
from sklearn.feature_selection import SelectKBest
from sklearn.impute import SimpleImputer
from sklearn.linear_model import LogisticRegression
from sklearn2pmml.decoration import ContinuousDomain
from sklearn2pmml.pipeline import PMMLPipeline
pipeline = PMMLPipeline([
("mapper", DataFrameMapper([
(["Sepal.Length", "Sepal.Width", "Petal.Length", "Petal.Width"], [ContinuousDomain(), SimpleImputer()])
])),
("pca", PCA(n_components = 3)),
("selector", SelectKBest(k = 2)),
("classifier", LogisticRegression(multi_class = "ovr"))
])
pipeline.fit(iris_X, iris_y)
pipeline.verify(iris_X.sample(n = 15))
from sklearn2pmml import sklearn2pmml
sklearn2pmml(pipeline, "LogisticRegressionIris.pmml", with_repr = True)