[FIXED] What does the target mean in fetch_20newsgroups

Issue

Sorry for asking dumb question. I'm new to the fetch_20newsgroup. I have printed out the newsgroups_train.target but i still don't get it what it means. Also, I have taken a look at the doc of sklearn. It said that target is a target label? Please someone tell me what does it mean thanks!

print(newsgroups_train.target)
[7 4 4 ... 3 1 8]

Solution

newsgroups_train.target returns the label corresponding to the features. It represents the ids of the newsgroup your are aiming to predict.

You can convert them to their respective names using newsgroups_train.target_names as follows :

from sklearn.datasets import fetch_20newsgroups
import numpy as np

newsgroups_train = fetch_20newsgroups(subset='train')

X, y = newsgroups_train.data, newsgroups_train.target

print(np.array(newsgroups_train.target_names)[y])

which output:

['rec.autos' 'comp.sys.mac.hardware' 'comp.sys.mac.hardware' ...
 'comp.sys.ibm.pc.hardware' 'comp.graphics' 'rec.motorcycles']

Answered By - Antoine Dubuis