我有一个csv数据，数据的第一列是'label'，第一列到第784列之后的列包含一个图像（28 * 28）格式的表示。我创建了一个numpy数组的元组使用load（filename）函数。此功能为图像指定标签。
现在我想使用下面函数read（digits，path =“。”）中生成的数据。此格式用于scikit库的支持向量机分析。事实上，我试图模仿（第二个例子是底部）：http://cvxopt.org/applications/svm/index.html
当我尝试重新格式化数据时，我收到了正文中下面复制的错误。有没有办法，我可以用所需的格式获得它？
train_name=dir_path+'train8.csv'
def load(filename):
# read file into a list of rows
with open(filename, 'rU') as csvfile:
lines = csv.reader(csvfile, delimiter=',')
rows = list(lines)
# create empty numpy arrays of the required size
data = np.empty((len(rows), len(rows[0])-1), dtype=np.float64)
expected = np.empty((len(rows),), dtype=np.int64)
# fill array with data from the csv-rows
for i, row in enumerate(rows):
data[i,:] = row[1:]
expected[i] = row[0]
result_data = data, expected
return result_data
> Result:
(array([[ 0., 0., 0., ..., 0., 0., 0.],
[ 0., 0., 0., ..., 0., 0., 0.],
[ 0., 0., 0., ..., 0., 0., 0.],
...,
[ 0., 0., 0., ..., 0., 0., 0.],
[ 0., 0., 0., ..., 0., 0., 0.],
[ 0., 0., 0., ..., 0., 0., 0.]]), array([1, 1, 1, ..., 1, 1, 1]))
def read(digits, path = "."):
data= load(train_name)
print "sizeImages = ",len(data[0]), "sizelabels = ", len(data[1])
lbl=data[1]
size=len(data[1])
img=data[0] #print type(img) OUTPUT: <type 'numpy.ndarray'>
img =[l[0] for l in img] # print type(img) OUTPUT: <type 'list'>,
*This is used to unpack the numpy array from above.*
ind = [ k for k in xrange(size) if lbl[k] in digits ]
images = matrix(0, (len(ind), 28*28))
labels = matrix(0, (len(ind), 1))
#images = data[0]
labels = img
for i in xrange(len(ind)):
images[i, :] = img[ ind[i]*28*28: (ind[i]+1)*28*28]
labels[i] = lbl[ind[i]]
return images,labels
print read([8], path = dir_path)
结果：
sizeImages = 5851 sizelabels = 5851
Traceback (most recent call last):
File "svm.py", line 62, in <module>
print read([8], path = dir_path)
File "svm.py", line 59, in read
images[i, :] = img[ ind[i]*28*28: (ind[i]+1)*28*28]
NotImplementedError: invalid type in assignment
所需格式：
**(<5949x784 matrix, tc='i'>, <5949x1 matrix, tc='i'>)**
也就是说，上面的第一个矩阵是：{array-like，sparse matrix}，shape = [n_samples，n_features]训练向量，其中样本数和n_features中的n_samples是特征数。
sample data:
'''1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 38 254 109 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 87 252 82 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 135 241 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 45 244 150 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 84 254 63 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 202 223 11 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 32 254 216 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 95 254 195 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 140 254 77 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 57 237 205 8 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 124 255 165 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 171 254 81 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 24 232 215 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 120 254 159 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 151 254 142 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 228 254 66 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 61 251 254 66 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 141 254 205 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 10 215 254 121 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 198 176 10 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0'''
答案 0 :(得分：0)
一个可能的解决方案是：
from sklearn.datasets.base import Bunch
file=np.loadtxt('~/data/test3.txt', dtype=int)
target_index=0
dataset = file
data = None
target = None
target_names = None
feature_names = None
# Target assumed to be either last or first row
if target_index == -1:
data = dataset[:, 0:-1]
target = dataset[:, -1]
elif target_index == 0:
data = dataset[:, 1:]
target = dataset[:, 0]
else:
raise ValueError("Target index must be either -1 or 0")
reformatData=Bunch(data=data, target=target)
print reformatData['data'].shape, reformatData['target'].shape