Python 和 scikit learn：我的模型的输出是 nan 而不是类别名称

Question

我正在研究基于输入预测输出类别的 ML 模型。我有一个没有错误的工作模型，但是，我将 nan 作为输出而不是“类别”值。我正在处理的数据都是文本。这是我的代码：import pandas as pdimport numpy as npdf=pd.read_excel('D:\\android\\medicare.xlsx')X=df['Product Description'].fillna(' ')Y=df['Category'].astype(str)from sklearn.model_selection import train_test_splitX_train,X_test,Y_train,Y_test=train_test_split(X,Y,test_size=.25,random_state=42)from sklearn.feature_extraction.text import CountVectorizercount_vector=CountVectorizer()X_train_count=count_vector.fit_transform((X_train).values.astype('U'))from sklearn.feature_extraction.text import TfidfTransformertfidf_transformer= TfidfTransformer()X_train_tfidf=tfidf_transformer.fit_transform(X_train_count)X_train_tfidf.shapefrom sklearn.naive_bayes import MultinomialNBclf = MultinomialNB().fit(X_train_tfidf, Y_train)from sklearn.pipeline import Pipelinefrom sklearn.externals import joblibimport pickletext_clf=Pipeline([('vect',CountVectorizer()),('tfidf',TfidfTransformer()),('clf',MultinomialNB()),])text_clf=text_clf.fit(X_train,Y_train)joblib.dump(text_clf,'model.pkl')X_test1=['SOTALOL 160MG CP SEC']predicted=text_clf.predict(X_test1)proab=text_clf.predict_proba(X_test1)print (str(predicted[0]))print (proab)print (text_clf.classes_)print (max(proab[0]))这是我的输出：我期待一个类别代码，但输出为“nan”。nan[[3.79853900e-06 2.84302863e-05 7.59252188e-06 ... 2.84280220e-05  1.89960087e-06 4.28977861e-04]]['153 Sm-SAMARIUM ACIDE ETHYLENEDIAMINETETRAMETHYLENE PHOSPHONIQUE' 'ABAISSE LANGUE' 'ABATACEPT' ... 'solutions salines' 'Électrodes ou câbles pour endoscopie' 'Étiquettes médicales à usage général ']0.8404466876175863