Index Distribution

How is the distribution of category records within the data matrix?

It's crucial to ascertain if the data entries linked to a specific category are randomly scattered across row indices or if they exhibit any form of clustering.

Oppgave

Swipe to start coding

Plot histograms for the distribution of indices for each digit (0-9) in the MNIST dataset. These histograms should be displayed on the individual subplots of the grid created.

Løsning

import numpy as np

# Create a 2x5 subplot grid, specify figure size, enable tight layout, and share the y-axis across subplots

f, ax = plt.subplots(2, 5, figsize=(12, 5), tight_layout=True, sharey=True)

# Iterate through digits 0-9 to plot their indices' distribution histograms on the subplots

for digit in range(10):

# Configure subplot title and xlabel; plot histogram for each digit's occurrence

sns.histplot(np.where(Y == digit)[0], bins=30, ax=ax[digit // 5, digit % 5], color='b').set(title=digit, xlabel='index')

Mark tasks as Completed

Alt var klart?

Takk for tilbakemeldingene dine!

Seksjon 1. Kapittel 5

AVAILABLE TO ULTIMATE ONLY