[PYTHON] from sklearn.datasets import fetch_mldata-> fetch_openml

background

An event that occurred when I was trying to cluster with t-SNE and was fishing various sample codes I was coding with reference to this

• https://towardsdatascience.com/visualising-high-dimensional-datasets-using-pca-and-t-sne-in-python-8ef87e7915b

Location of occurrence

Part 1

import matplotlib.pyplot as plt
from mpl.toolkits.mplot3d import Axes3D
import seaborn as sns
from __future__ import print_function
from sklearn.datasets import fetch_openml # <-here
from sklearn.decompositon import PCA
from sklearn.manifold import TSNE

%matplotlib inline

Part 2

mnist = fetch_mldata("MNIST original") # <-here
X = mnist.data / 255.0
y = mnist.target
print(X.shape, y.shape)

Correction code

Part 1

import matplotlib.pyplot as plt
from mpl.toolkits.mplot3d import Axes3D
import seaborn as sns
from __future__ import print_function
from sklearn.datasets import fetch_openml # <-here
from sklearn.decompositon import PCA
from sklearn.manifold import TSNE

%matplotlib inline

Part 2

mnist = fetch_openml('mnist_784', version=1,) # <-here
X = mnist.data / 255.0
y = mnist.target

print(X.shape, y.shape)

Cause

It seems that the cause is that https://openml.org, which is the reference destination of fetch_mldata, is down.

reference

• https://github.com/ageron/handson-ml/issues/529
• https://github.com/scikit-learn/scikit-learn/issues/8588#issuecomment-292634781
• https://github.com/YutaroOgawa/Deep-Reinforcement-Learning-Book/issues/4 This is Japanese