Function reference • littleDeep

All functions
`ImageArray-class`	container for 4D image array (index,x,y,z), typically (,32,32,3)
`ImageArray()`	constructor for ImageArray
`SCE2ImageArray()`	convert assay data from SCE to ImageArray
`build_array()`	make an array of JPEG-like data for invocations of a (presumably random) shape-generating function
`chkthres()`	app to demonstrate filtering of a jpeg
`cifar100_history`	a history object for keras fit for cifar100 data, using the ISLR tuning settings
`eval_model()`	estimate accuracy of fitted model predictions for a given ImageArray
`filterByType()`	return an ImageArray with images restricted to specific types
`flattenToMatrix()`	linearize data in an image to a single vector for each image in an ImageArray instance
`getArray(<ImageArray>)`	extract numerical array
`getArray()`	extract numerical array
`getTypes(<ImageArray>)`	extract vector of labels
`getTypes()`	extract vector of labels
`get_fitted_cifar100()`	restore a fitted model for CIFAR100
`getline()`	produce sequence of points on a line between two points
`islr_cnn()`	use the CNN in Introduction to Statistical Learning with R on an ImageArray instance
`jpeg_shrinker()`	app for uploading and transforming a square jpeg
`load_jpeg()`	this function defines a 3-dimensional array with given dimensions and sets all elements to 1 except those with element indices given by x and y, which are set to zero
`load_shape_cnn()`	load shape-trained CNN
`make_shape_iarr()`	produce an ImageArray of circles, triangles, quadrilaterals
`makecirc()`	make a discrete image of a circle
`makequad()`	produce a list of x and y coordinates of "discrete lines" with specified numbers of points, forming a quadrilateral with vertices given by input coordinates
`maketri()`	produce a list of x and y coordinates of "discrete lines" with specified numbers of points, forming a triangle with vertices given by input coordinates
`model_probs()`	report estimated model properties for a given ImageArray
`preview()`	produce labeled 3x3 array of images via plot.raster
`print(<islr_cnn>)`	print summary of islr_cnn object
`process_jpg()`	import, resize and rotate a JPEG that should have been produced in square mode
`ranbox()`	crude approach to a random rectangle in 32x32 rectangle, can bleed over
`rancirc()`	produce coordinates for a randomly positioned circle in a square plane
`ranquad()`	produce coordinates of a random "quadrilateral" in 64 x 64 plane
`rantri()`	produce coordinates of a random "triangle" in 64 x 64 plane
`restore_islr_cnn()`	restore fitted model and addons
`run_cifar100()`	run the ISLR CNN for CIFAR100 data
`save_islr_cnn()`	save a fitted model in a folder with hdf5 and list components
`show(<ImageArray>)`	display information about ImageArray instance
`show_jpg()`	produce raster image of a jpg file
`[`(<ImageArray>,<numeric>,<missing>,<missing>)	simple extraction with bracket
`train_cifar100()`	build islr_cnn instance trained on cifar100

Reference

All functions