这似乎是我见过的常见错误,但有一些潜在的原因。SciKit-Learn:预测时出错
我想在Python中放在一起逻辑回归。我的数据被保存为熊猫数据框。
train, test = train_test_split(final_dat[train_cols], train_size=0.80, random_state=1)
logit = sm.Logit(train['SPR_Created__c'], train.drop(['SPR_Created__c'], axis=1))
result = logit.fit()
print result.summary()
result.predict(test[train_cols])
错误:
result.predict(test[train_cols])
ValueError: shapes (13664,18) and (17,) not aligned: 18 (dim 1) != 17 (dim 0)
我不知道这个错误occurence所有最变量进行了调整。
final_dat[train_cols].info()
<class 'pandas.core.frame.DataFrame'>
Int64Index: 68319 entries, 0 to 31978
Data columns (total 18 columns):
Is_Subject 68319 non-null int64
Is_Description 68319 non-null int64
SPR_Created__c 68319 non-null int64
description2_contains_any_synonym 68319 non-null int64
description_length 68319 non-null int64
subject_length 68319 non-null int64
description2_length 68319 non-null int64
Is_Description2 68319 non-null int64
Is_Adverse_Event 68319 non-null int64
subject_contains_common_spr_terms 68319 non-null int64
description_contains_common_spr_terms 68319 non-null int64
description2_contains_common_spr_terms 68319 non-null int64
pattern_exists_in_description 68319 non-null int64
pattern_exists_in_description_count 68319 non-null float64
pattern_exists_in_description2 68319 non-null int64
pattern_exists_in_description2_count 68319 non-null float64
subject_contains_any_synonym 68319 non-null int64
description_contains_any_synonym 68319 non-null int64
dtypes: float64(2), int64(16)
memory usage: 12.4 MB
关于什么可能是错误的任何想法?
您可以将您的数据? – sera