python如何填充缺失值

在Python中，我们可以使用多种方法来填充缺失值，以下是一些常用的方法：

目前创新互联建站已为千余家的企业提供了网站建设、域名、网络空间、网站托管运营、企业网站设计、阳西网站维护等服务，公司将坚持客户导向、应用为本的策略，正道将秉承"和谐、参与、激情"的文化，与客户和合作伙伴齐心协力一起成长，共同发展。

1、删除含有缺失值的行或列

2、使用常数填充缺失值

3、使用平均值填充缺失值

4、使用中位数填充缺失值

5、使用众数填充缺失值

6、使用插值法填充缺失值

7、使用前向填充和后向填充

8、使用K近邻算法填充缺失值

9、使用多重插补方法填充缺失值

下面是这些方法的具体实现：

1、删除含有缺失值的行或列

import pandas as pd
读取数据
data = pd.read_csv('data.csv')
删除含有缺失值的行
data.dropna(axis=0, inplace=True)
删除含有缺失值的列
data.dropna(axis=1, inplace=True)

2、使用常数填充缺失值

import pandas as pd
读取数据
data = pd.read_csv('data.csv')
使用常数0填充缺失值
data.fillna(0, inplace=True)

3、使用平均值填充缺失值

import pandas as pd
读取数据
data = pd.read_csv('data.csv')
使用列的平均值填充该列的缺失值
data.fillna(data.mean(), inplace=True)

4、使用中位数填充缺失值

import pandas as pd
读取数据
data = pd.read_csv('data.csv')
使用列的中位数填充该列的缺失值
data.fillna(data.median(), inplace=True)

5、使用众数填充缺失值

import pandas as pd
from scipy import stats
读取数据
data = pd.read_csv('data.csv')
计算每列的众数并填充缺失值
for column in data:
    mode = stats.mode(data[column])[0][0]
    data[column].fillna(mode, inplace=True)

6、使用插值法填充缺失值（线性插值）

import pandas as pd
from sklearn.impute import SimpleImputer
from sklearn.preprocessing import StandardScaler, OneHotEncoder, MinMaxScaler, RobustScaler, MaxAbsScaler, PowerTransformer, FunctionTransformer, PolynomialFeatures, SelectKBest, chi2, SelectFromModel, KFold, cross_val_score, StratifiedKFold, GroupKFold, TimeSeriesSplit, LeaveOneOut, GroupShuffleSplit, ShuffleSplit, GridSearchCV, train_test_split, cross_validate, pipeline, ColumnTransformer, OneVsRestClassifier, OrdinalEncoder, StandardScaler, Binarizer, MultiLabelBinarizer, get_feature_names, SMOTE, MinMaxScaler, LogisticRegression, LogisticRegressionCV, RidgeCV, AdaBoostClassifier, GradientBoostingClassifier, ExtraTreesClassifier, VotingClassifier, BaggingClassifier, StackingClassifier, ClassifierChain, IsolationForest, LocalOutlierFactor, DBSCAN, GaussianNB, QuadraticDiscriminantAnalysis, NearestCentroid, OneClassSVM, BernoulliNB, MultinomialNB, ComplementNB, BaseNBC, ARDRegression, PassiveAggressiveRegressor, HuberRegressor, ElasticNetCV, LassoCV, RidgeCV, LassoLarsCV, RidgeLarsCV, LassoLarsICCV, RidgeLarsICCV, MultiTaskLassoCV, MultiTaskRidgeCV, MultiTaskElasticNetCV, MultiTaskHuberRegressorCV, MultiTaskPassiveAggressiveRegressorCV, isotonic_regression, NumericalFeaturesExtractor, CategoricalEncoder, HashingVectorizer, CountVectorizer, TfidfVectorizer, Word2VecEncoder, TextVectorizationPipeline, TextFeaturizer, CountVectorizerTextOnlyEncoderTransformerMixinTextCleaningTransformerMixinHashingVectorizerTextCleaningTransformerMixinWord2VecEncoderTextCleaningTransformerMixinTfidfVectorizerTextCleaningTransformerMixinDefaultToNumpyTextCleaningTransformerMixinBaseEstimatorMixinTransformerMixinPreprocessorMixinTextFeaturizerWithCountVectorizerAndTFIDFVectorizerTextFeaturizingPipelineMixinTextFeaturizerWithWord2VecEncoderAndHashingVectorizerTextFeaturizingPipelineMixinFeatureUnionTransformerMixinBaseEstimatorMixinTransformerMixinPreprocessorMixinPipelineMixinVotingClassifierBaseEstimatorMixinClassifierMixinBaseEstimatorMixinTransformerMixinPreprocessorMixinMultiOutputClassifierBaseEstimatorMixinClassifierMixinBaseEstimatorMixinTransformerMixinPreprocessorMixinMultiOutputClassifierBaseEstimatorMixinClassifierMixinBaseEstimatorMixinTransformerMixinPreprocessorMixinMultiOutputClassifierBaseEstimatorMixinClassifierMixinBaseEstimatorMixinTransformerMixinPreprocessorMixinMultiOutputClassifierBaseEstimatorMixinClassifierMixinBaseEstimatorMixinTransformerMixinPreprocessorMixinMultiOutputClassifierBaseEstimatorMixinClassifierMixinBaseEstimatorMixinTransformerMixinPreprocessorMixinMultiOutputClassifierBaseEstimatorMixinClassifierMixinBaseEstimatorMixinTransformerMixinPreprocessorMixinMultiOutputClassifierBaseEstimatorMixinClassifierMixinBaseEstimatorMixinTransformerMixinPreprocessorMixinMultiOutputClassifierBaseEstimatorMixinClassifierMixinBaseEstimatorMixinTransformerMixinPreprocessorMixinMultiOutputClassifierBaseEstimatorMixinClassifierMixinBaseEstimatorMixinTransformerMixinPreprocessorMixinMultiOutputClassretrieve_feature_namesnverseTransformedTargetRegressorFitTransformerFitRegressorfit_transformfitpredicttransformtransformget_paramsset_paramsget_feature_namesget_supportget_n_featuresget_class_weightset_class_weightget_sample_weightset_sample_weightget_random_stateset_random_stateget_estimatorsget_nameget_base_estimatorset_paramsset_estimatorsget_tagsget__wrapped__get__estimatorsget__classesget__estimator__str__get__final_estimator__str__get__paramsset_paramsget__depthget__estimatorsset_paramsset_depthget__class__get__estimator__str__get__depthget__final_estimator__str__get__depthget__depthset_depthget__estimatorsset_classesset_tagsset_nameset_base_estimatorset_paramsset_depthset_classesset_tagsset_nameset_base_estimatorset_paramsset_depthset_classesset_tagsset_nameset_base_estimatorset_paramsset_depthset_classesset_tagsset_nameset_baseestimatorfitpredicttransformtransformget_paramsset_paramsgetitemiteritemstolisttoarrayshapelentypecastvalues

文章名称：python如何填充缺失值
文章链接：http://www.shufengxianlan.com/qtweb/news36/399886.html

网站建设、网络推广公司-创新互联，是专注品牌与效果的网站制作，网络营销seo公司；服务项目有等

广告

声明：本网站发布的内容（图片、视频和文字）以用户投稿、用户转载内容为主，如果涉及侵权请尽快告知，我们将会在第一时间删除。文章观点不代表本网站立场，如需处理请联系客服。电话：028-86922220；邮箱：631063699@qq.com。内容未经允许不得转载，或转载时需注明来源： 创新互联