我有一组颜色列表:
initialColors = [u'black' u'black' u'black' u'white' u'white' u'white' u'powderblue' u'whitesmoke' u'black' u'cornflowerblue' u'powderblue' u'powderblue' u'goldenrod' u'white' u'lavender' u'white' u'powderblue' u'powderblue' u'powderblue' u'powderblue' u'powderblue' u'powderblue' u'powderblue' u'powderblue' u'white' u'white' u'powderblue' u'white' u'white']
这些颜色有对应的标签如下:
labels_train = [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1]
0
表示颜色由女性选择,1
表示由男性选择。我打算使用另一组颜色来预测性别。
因此,对于我的初始颜色,我将颜色名称转换为数值特征向量,如下所示:
from sklearn import preprocessingle = preprocessing.LabelEncoder()le.fit(initialColors)features_train = le.transform(initialColors)
转换后,我的features_train
看起来像这样:
[0 0 0 5 5 5 4 6 0 1 4 4 2 5 3 5 4 4 4 4 4 4 4 4 5 5 4 5 5]
最后,我执行以下操作:
from sklearn.naive_bayes import GaussianNBclf = GaussianNB()clf.fit(features_train, labels_train)
但我得到了一个错误:
/Library/Python/2.7/site-packages/sklearn/utils/validation.py:395: DeprecationWarning: Passing 1d arrays as data is deprecated in 0.17 and will raise ValueError in 0.19. Reshape your data either using X.reshape(-1, 1) if your data has a single feature or X.reshape(1, -1) if it contains a single sample. DeprecationWarning)Traceback (most recent call last): File "app.py", line 36, in <module> clf.fit(features_train, labels_train) File "/Library/Python/2.7/site-packages/sklearn/naive_bayes.py", line 182, in fit X, y = check_X_y(X, y) File "/Library/Python/2.7/site-packages/sklearn/utils/validation.py", line 531, in check_X_y check_consistent_length(X, y) File "/Library/Python/2.7/site-packages/sklearn/utils/validation.py", line 181, in check_consistent_length " samples: %r" % [int(l) for l in lengths])ValueError: Found input variables with inconsistent numbers of samples: [1, 70]
我尝试了以下操作:
features_train = features_train.reshape(-1, 1)labels_train = labels_train.reshape(-1, 1)clf.fit(features_train, labels_train)
我得到了一个错误:
/Library/Python/2.7/site-packages/sklearn/utils/validation.py:526: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel(). y = column_or_1d(y, warn=True)
我也尝试了:
features_train = features_train.reshape(1, -1)labels_train = labels_train.reshape(1, -1)
但无论如何:
Traceback (most recent call last): File "app.py", line 36, in <module> clf.fit(features_train, labels_train) File "/Library/Python/2.7/site-packages/sklearn/naive_bayes.py", line 182, in fit X, y = check_X_y(X, y) File "/Library/Python/2.7/site-packages/sklearn/utils/validation.py", line 526, in check_X_y y = column_or_1d(y, warn=True) File "/Library/Python/2.7/site-packages/sklearn/utils/validation.py", line 562, in column_or_1d raise ValueError("bad input shape {0}".format(shape))ValueError: bad input shape (1, 29)
我的问题是我不明白在我的情况下重塑数据的最佳方法是什么。你能帮我选择一种重塑数据的方法吗?
回答:
快速回答:
- 执行
features_train = features_train.reshape(-1, 1)
; - 不要执行
labels_train = labels_train.reshape(-1, 1)
。保持labels_train
不变。
一些细节:
看起来你对为什么估计器需要2D数据数组输入感到困惑。你的训练向量X
具有形状(n_samples, n_features)。所以features_train.reshape(-1, 1)
在你的情况下是正确的,因为你只有一个特征,并且希望让numpy
推断出有多少样本。这确实解决了你的第一个错误。
你的目标值y
具有形状(n_samples,),它期望一个1D数组。当你执行labels_train = labels_train.reshape(-1, 1)
时,你将其转换为2D列向量。这就是你得到第二个警告的原因。请注意,这是一个警告,意味着fit()
已经弄明白了并进行了正确的转换,即你的程序继续运行并且应该是正确的。
当你执行以下操作时:
features_train = features_train.reshape(1, -1)labels_train = labels_train.reshape(1, -1)
首先,对于你的情况来说,这对features_train
来说是一个错误的转换,因为X.reshape(1, -1)
意味着你有一个样本,并且希望让numpy
推断出有多少特征。这不是你想要的,但fit()
不知道,会相应地处理它,给你错误的结果。
话虽如此,你的最后一个错误并不是来自features_train = features_train.reshape(1, -1)
。它来自labels_train = labels_train.reshape(1, -1)
。你的labels_train
现在具有形状(1, 29),这既不是行向量也不是列向量。虽然我们可能知道它应该被解释为目标值的1D数组,但fit()
还没有那么聪明,不知道该怎么处理它。