ploterrhist

(To be removed) Plot error histogram

collapse all in page

ploterrhist will be removed in a future release. For more information, see Transition Legacy Neural Network Code to dlnetwork Workflows.

For advice on updating your code, see Version History.

Syntax

ploterrhist(e)
ploterrhist(e1,'name1',e2,'name2',...)
ploterrhist(...,'bins',bins)

Description

ploterrhist(e) plots a histogram of error values e.

ploterrhist(e1,'name1',e2,'name2',...) takes any number of errors and names and plots each pair.

ploterrhist(...,'bins',bins) takes an optional property name-value pair which defines the number of bins to use in the histogram plot. The default is 20.

example

Examples

collapse all

This example shows how to plot the histogram of error values of a trained feedforward network.

Create a feedforward network and train it using the data from the simple fit data set. Then plot the histogram of error values.

[x,t] = simplefit_dataset;
net = feedforwardnet(20);
net = train(net,x,t);
y = net(x);
e = t - y;
ploterrhist(e,'bins',30)

Histogram of error values. The errors are clustered around zero in a roughly Gaussian distribution.

Input Arguments

collapse all

Errors to plot, specified as a vector or a matrix.

Number of bins to use in the histogram plot, specified as the comma-separated pair consisting of 'bins' and an integer.

Version History

Introduced in R2010b

expand all

See Also

| (Statistics and Machine Learning Toolbox) | | | |

Topics