Question

# Extract the features for a given application
features = extract_features(file_path)
# Form the feature vector
feature_vector = create_vector_single(features)
# Load the pre configured feature model from a pickle file
model = pickle.load(open("feature_model.p", "rb"))
# Reduce the feature vector into size 12
feature_vector_new = model.transform(feature_vector)

# Load the pre-trained model from a pickle file
clf = pickle.load( open( "kfold_train_data.p", "rb" ) )

# Perform prediction using the model
result = clf.predict(feature_vector_new)

feature_vector_new = model.transform(feature_vector)上发生错误

Reshape your data either using array.reshape(-1, 1) if your data has a single feature or array.reshape(1, -1) if it contains a single sample.

然后我用：

解决

# Reduce the feature vector into size 12
vec = [feature_vector]
vec = np.array(vec).reshape(1, -1)
feature_vector_new = model.transform(vec)

它会生成另一个错误：

AttributeError: 'SelectFromModel' object has no attribute 'norm_order'

Answer 1

当我跟踪类似的代码时，我发现生成的模型对象中的参数DELETE FROM tblParts WHERE NOT EXISTS ( SELECT '1' FROM tblOrders WHERE tblOrders.PartID = tblParts.ID );少了一个norm_order。

答案：重新生成一个模型，然后这个问题就解决了！

```

feature_model.p

```

关于机器学习：使用array.reshape（-1，1）重塑数据

1 个答案: