inst/tmp.R

In matloff/imageClassR: Quick and Easy Functions for Image Classification

## source("TDAsweep.R")
## 
## # Read data
## mnist <- read.csv('../mnist.csv')
## 
## # Make data into `images` matrix and `labels` vector
## # images <- matrix(unlist(mnist[, 1:784]), ncol=784, byrow=TRUE)
## images <- as.matrix(mnist[, 1:784])
## labels <- as.matrix(mnist[785])
## # found this in https://stackoverflow.com/a/13224720
## 
## # Plot data
## for (i in 1:70000) {
## 	filled.contour(matrix(images[i, ], 28, byrow=T))
## }
## 
## # Invoke TDAsweepOneImg?
## TDAsweepOneImg(matrix(images[1,], ncol=28), 28, 28)