如何解决 StandardScaler 值错误答案

【问题标题】：how to resolve StandardScaler value error如何解决 StandardScaler 值错误
【发布时间】：2022-01-09 09:34:48
【问题描述】：

我正在按照image to image search 教程对图像数据集执行语义搜索。我搜索了多个 SO 问题和答案，但可以查明解决方案。这是我的代码

path="../../datasets/101_ObjectCategories"
dest = "content/all_images/"

all_images = [dest+elem for elem in os.listdir(dest)]    #2
print("AMOUNT OF IMAGES :", len(all_images))

# Load as numpy array
dataset = np.empty(shape=(len(all_images),128,128,3)) #1
# print("DATASETS :", dataset)

for i,path in enumerate(all_images):  #2
    img = load_img(path, target_size=(128,128))
    img_arr = img_to_array(img)
    dataset[i] = img_arr

np.random.shuffle(dataset) #3

dshape = dataset.shape
print("dataset shape ", dshape)


pipeline = Pipeline([('scaling', StandardScaler()), ('pca', PCA(n_components=5))])

pipeline.fit_transform(dataset)

我目前遇到错误

这是我的输出打印输出错误

dataset shape  (9171, 128, 128, 3)
Traceback (most recent call last):
  File "embedder1_main.py", line 78, in <module>
    pipeline.fit_transform(dataset)
  File "/Users/james/miniforge3/envs/konstant-3/lib/python3.8/site-packages/sklearn/pipeline.py", line 426, in fit_transform
    Xt = self._fit(X, y, **fit_params_steps)
  File "/Users/james/miniforge3/envs/konstant-3/lib/python3.8/site-packages/sklearn/pipeline.py", line 348, in _fit
    X, fitted_transformer = fit_transform_one_cached(
  File "/Users/james/miniforge3/envs/konstant-3/lib/python3.8/site-packages/joblib/memory.py", line 349, in __call__
    return self.func(*args, **kwargs)
  File "/Users/james/miniforge3/envs/konstant-3/lib/python3.8/site-packages/sklearn/pipeline.py", line 893, in _fit_transform_one
    res = transformer.fit_transform(X, y, **fit_params)
  File "/Users/james/miniforge3/envs/konstant-3/lib/python3.8/site-packages/sklearn/base.py", line 847, in fit_transform
    return self.fit(X, **fit_params).transform(X)
  File "/Users/james/miniforge3/envs/konstant-3/lib/python3.8/site-packages/sklearn/preprocessing/_data.py", line 806, in fit
    return self.partial_fit(X, y, sample_weight)
  File "/Users/james/miniforge3/envs/konstant-3/lib/python3.8/site-packages/sklearn/preprocessing/_data.py", line 841, in partial_fit
    X = self._validate_data(
  File "/Users/james/miniforge3/envs/konstant-3/lib/python3.8/site-packages/sklearn/base.py", line 561, in _validate_data
    X = check_array(X, **check_params)
  File "/Users/james/miniforge3/envs/konstant-3/lib/python3.8/site-packages/sklearn/utils/validation.py", line 786, in check_array
    raise ValueError(
ValueError: Found array with dim 4. StandardScaler expected <= 2.

是什么导致了这个错误，我该如何解决？

【问题讨论】：

标签： python numpy scikit-learn

【解决方案1】：

如打印输出所述，您的数据集形状为 (9171, 128, 128, 3) 所以 dim = 4。

您提供的链接中的代码使用深度神经网络来训练模型，让作者有机会轻松处理多波段图像 (RGB)。此外，为了处理 0 到 1 之间的值，作者将图像数组除以 255（最大可能值）。

如果您仍然想使用StandardScaler，恐怕您需要一个自定义函数来分别处理每个图像和每个波段，以便每次都能处理一个(128,128) 数组，虽然我是不确定是最好的方法。

【讨论】：