README.md
In jlmelville/snedata: SNE Simulation Dataset Functions

snedata

SNE Dataset Functions for R

This package provides functions for generating simple simulation datasets for use in Stochastic Neighbor Embedding and related dimensionality reduction methods, most obviously the very popular t-SNE.

November 25 2023 Added the 20 Newsgroups text dataset. You'll need to look at packages like tm or tidytext to process this into a form suitable for embedding.

May 14 2022 Added the taspheres function to generate the "spheres" dataset of high dimensional spheres nested inside a larger sphere, as used in the Topological Autoencoders paper.

December 24 2021. Added the S-curve with a hole and 2D curve datasets used in the PaCMAP paper.

December 21 2021. Added the mammoth datasets from Understanding UMAP.

June 29 2019. Added QMNIST.

March 6 2019. Added CIFAR-10.

February 23 2019. Added Small NORB.

December 14 2018. Added Kuzushiji-MNIST.

Synthetic data similar to those used by Lee and co-workers in their JSE and multi-JSE papers.
Also, Martin Wattenberg, Fernanda Viégas and Ian Johnson published an interactive article How to use t-SNE Effectively. The JavaScript functions used to create the simulation datasets (which can also be found at https://github.com/distillpub/post--misread-tsne), have been translated into R and are also hosted in this package.
Additionally, if you have the RnavGraphImageData package installed, there are also functions to convert the Olivetti and Frey faces datasets into a row-based data frame and functions to visualize them.
Code to download and visualize the MNIST database, based on a gist by Brendan O'Connor, who graciously allowed it to be MIT-licensed.
QMNIST extends the MNIST test dataset to 60,000 digits.
The Fashion-MNIST dataset is intended as drop-in replacement for the MNIST digits database, but using images of fashion items, and to be harder to perform well with machine learning benchmarks. Items can be visualized with the same function intended for the MNIST digits.
The Kuzushiji-MNIST dataset, another drop-in replacement for the MNIST digits, but this time of types of cursive Japanese characters.
The Small NORB dataset, which consists of pairs of images of 50 toys from different angles and under different lighting conditions.
The CIFAR-10 dataset, which consists of 60,000 32 x 32 color images in ten different classes.
A 3D point cloud of a mammoth at the Smithsonian, from Understanding UMAP, based on work originally done by Max Noichl.
A 3D S-curve with a hole data set, used to validate the PaCMAP method (see also the github repo). Bonus: a translation from Python of the sklearn.datasets.make_s_curve function. Also: a simple 2D curve dataset, useful for testing initialization strategies.

install.packages("remotes")
remotes::install_github("jlmelville/snedata")

package?snedata # lists all the functions
?snedata::gaussian_data # contains links to all the other distill.pub functions

Examples

library(snedata)

# 3000 points sampled from the surface of a sphere
sphere3000 <- sphere(n = 3000)

# 1500 points sampled from a toroidal helix with 30 coils:
helix1500 <- helix(n = 1500, nwinds = 30)

# 1500 points from a filled sphere:
ball1500 <- ball(n = 1500)

# 1000 points from a "Swiss Roll" distribution:
swiss1000 <- swiss_roll(n = 1000)

# 1000 points from a five-dimensional gaussian:
gauss1000 <- gaussian_data(n = 1000, dim = 5)

# Load RnavGraphImageData
library(RnavGraphImageData)

# Load the Frey faces dataset with each image as a row
frey <- frey_faces()
# Display the first pose
show_frey_face(frey, 1)

# PCA scores plot, with color indicating the frame index
frey_pca <- prcomp(frey[, -561], retx = TRUE, rank. = 2)
plot(frey_pca$x, col = frey$color, pch = 16, cex = 0.75)

# Load the Olivetti faces dataset with each image as a row
olivetti <- olivetti_faces()
# Show the second pose of the first face
show_olivetti_face(olivetti, 1, 2)

# Generate datasets similar to those used in the main text of "How to Use t-SNE Effectively"
misread_tsne <- list(
    two_clusters = two_clusters_data(n = 50, dim = 2),
    two_different_sized_clusters = two_different_clusters_data(n = 75, dim = 2),
    three_clusters_50 = three_clusters_data(n = 50, dim = 2),
    three_clusters_200 = three_clusters_data(n = 200, dim = 2),
    gaussian_cloud = gaussian_data(n = 500, dim = 100),
    ellipsoidal_gaussian_cloud = long_gaussian_data(n = 100, dim = 50),
    two_long_linear_clusters = long_cluster_data(n = 75),
    cluster_in_cluster = subset_clusters_data(n = 75, dim = 50),
    linked_rings = link_data(n = 100),
    trefoil_knot = trefoil_data(n = 150)
)

# fetch the MNIST data set from the MNIST website
mnist <- download_mnist()

# view the fifth digit
show_mnist_digit(mnist, 5)

# first 60,000 instances are the training set
mnist_train <- head(mnist, 60000)
# the remaining 10,000 are the test set
mnist_test <- tail(mnist, 10000)

# PCA on 1000 random training examples
mnist_r1000 <- mnist_train[sample(nrow(mnist_train), 1000), ]

pca <- prcomp(mnist_r1000[, 1:784], retx = TRUE, rank. = 2)
# plot the scores of the first two components
plot(pca$x[, 1:2], type = 'n')
text(pca$x[, 1:2], labels = mnist_r1000$Label, cex = 0.5,
  col = rainbow(length(levels(mnist_r1000$Label)))[mnist_r1000$Label])

# save to disk
save(mnist, file = "mnist.Rda")

# fetch the Fashion-MNIST dataset
fashion <- download_fashion_mnist()

# Works as a drop-in replacement for the MNIST digits, can repeat the above
# view the fifth item etc.
show_mnist_digit(fashion, 5)

# similarly for Kuzushiji-MNIST dataset of Japanese cursive characters
# (set verbose flag to see download progress)
kuzushiji <- download_kuzushiji_mnist(verbose = TRUE)
# View the tenth character
show_mnist_digit(kuzushiji, 10)

# Download the small NORB dataset
norb <- download_norb_small(verbose = TRUE)
# View an image, compare with example at https://github.com/ndrplz/small_norb
show_norb_object(norb, category = 2, instance = 6, elevation = 6, azimuth = 24, lighting = 2)