本文共 2445 字，大约阅读时间需要 8 分钟。

1、unexpected indent

Python对代码格式是很严格的，indentation是缩进、缩排的意思，unexpected indent 就是说有“意外的”缩进。这时，就要查看自己的代码格式了！

2、invalid syntax

invalid  adj.无效的；不能成立的；

syntax  n.语法；句法；句法规则[分析]；语构

invalid syntax就是无效语法。这时候，要检查自己的代码语法。

嘻嘻然后记入门学习NPL的今天的代码（还没搞懂代码的意思……）！

# -*- coding: utf-8 -*-"""Spyder EditorThis is a temporary script file.""" # -*- coding: utf-8 -*-  """  Created on Mon Aug 15 21:00:27 2016  @author: amnesia  """  	from sklearn.datasets import load_iris  iris = load_iris()  	# The feature (column) names and the response  print(iris.feature_names)print(iris.target)print(iris.target_names)	# The object types of the feature matrix and the response array  print(type(iris.data))print(type(iris.target))	# The shapes of samples and features  print(iris.data.shape)

# -*- coding: utf-8 -*-"""Created on Tue Aug 16 19:29:30 2016@author: amnesia"""from sklearn.datasets import load_iris# Import LinearSVC classfrom sklearn.svm import LinearSVC# Import KNeighborsClassifier classfrom sklearn.neighbors import KNeighborsClassifier# Load the datasetiris = load_iris()# Assign to variables for more convenient handlingX = iris.datay = iris.target# Create an instance of the LinearSVC classifierclf = LinearSVC()# Train the modelclf.fit(X, y)# Get the accuracy score of the LinearSVC classifierprint (clf.score(X, y))# Predict the response given a new observationprint (clf.predict([[ 6.3,  3.3,  6.0,  2.5]]))# Create an instance of KNeighborsClassifier# The default number of K neighbors is 5.# This can be changed by passing n_neighbors=k as argumentknnDefault = KNeighborsClassifier() # K = 5# Train the modelknnDefault.fit(X, y)# Get the accuracy score of KNeighborsClassifier with K = 5print (knnDefault.score(X, y))# Predict the response given a new observationprint (knnDefault.predict([[ 6.3,  3.3,  6.0,  2.5]]))# Let's try a different number of neighborsknnBest = KNeighborsClassifier(n_neighbors=10) # K = 10# Train the modelknnBest.fit(X, y)# Get the accuracy score of KNeighborsClassifier with K = 10print (knnBest.score(X, y))# Predict the response given a new observationprint (knnBest.predict([[ 6.3,  3.3,  6.0,  2.5]]))# Let's try a different number of neighborsknnWorst = KNeighborsClassifier(n_neighbors=100) # K = 100# Train the modelknnWorst.fit(X, y)# Get the accuracy score of KNeighborsClassifier with K = 100print (knnWorst.score(X, y))# Predict the response given a new observationprint (knnWorst.predict([[ 6.3,  3.3,  6.0,  2.5]]))

转载地址：http://ittai.baihongyu.com/