mnist: MNIST handwritten digits dataset
In xrobin/mnist: MNIST dataset

Description Usage Arguments Format Details Source Examples

A dataset of handwritten digits provided by http://yann.lecun.com/exdb/mnist/.

1 2	mnist download.mnist(range = c(0, 1), global = FALSE)

`range`	the range of the pixel intensity values. If `c(0, 255)` the pixel intensities will be the raw integer values. By default, they are continuous between 0 and 1. If possible an integer matrix is returned.
`global`	if `TRUE`, the `mnist` dataset will be permanently replaced in the package's directory and global environment. Useful to change the range in a persisting manner.

A list with:

train, test: 60000 train and 10000 test digits. Both contain x (data) and y (labels) elements. x is a matrix with 784 columns (pixels).
show.digit: a function to display a digit. Takes the following arguments:
- digit: a digit, i.e. a row of mnist$train$x or mnist$test$x
- col: a colour gradient, see image for more details
- ...: further arguments for image

At the first time the package is loaded, the mnist dataset is downloaded from http://yann.lecun.com/exdb/mnist/ and saved in the package data directory.

Y. LeCun, L. Bottou, Y. Bengio, and P. Haffner. "Gradient-based learning applied to document recognition." Proceedings of the IEEE, 86(11):2278-2324, November 1998.

The dataset is downloaded with code inspired from: https://gist.github.com/brendano/39760

data(mnist) # the dataset is implicitly loaded when the package is attached, so this line is optionnal
str(mnist)

show.digit(mnist$test$x[1:10,])

## Not run: 
mnist.copy <- download.mnist()

## End(Not run)