MNIST 随机样本绘图不起作用

Question

我正在随机抽取 MNIST data set，但它显示错误。我哪里做错了？

import sklearn
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
from sklearn.datasets import fetch_openml

mnist = fetch_openml('mnist_784')
y = mnist.target
X = mnist.data.astype('float64')

fig, ax = plt.subplots(2, 5)
ax = ax.flatten()
for i in range(10):
    im_idx = np.argwhere(y == str(i))[0]    # Pandas versions bug - it worked in PyCharm. {Kindly, what is an alternate way to fix it?}
    print(im_idx)
    plottable_image = np.reshape(X[im_idx], (28, 28))
    ax[i].imshow(plottable_image, cmap='gray_r')

错误显示：

ValueError: Length of passed values is 1, index implies 70000.

Answer 1

你得到结果的错误是因为 np.argwhere(y == i) returns 是一个空数组，那是因为你试图在 y 和 [=14= 填充的数组之间进行比较] 值和 i 这是一个 int.

以下更改将修复它：

im_idx = np.argwhere(y == str(i))[0]

这是匹配 NumPy 操作的转换：

y = mnist.target.to_numpy()
X = mnist.data.astype('float64').to_numpy()

MNIST 随机样本绘图不起作用

MNIST random sample plotting is not working

python

machine-learning

matplotlib

scikit-learn

mnist