CStats:Mean

CStats:Mean

The Mean method computes the mean and standard deviation of numeric data. The data may be in the form of a lua array or a class object of type CImage, CArray, or CMatrix. For the CImage and CMatrix classes, you can optionally specify a rectangle that bounds the data region. For a CImage, the data may have multiple channels as with an RGB image.

Syntax

nMean, nSdev = CStats:Mean( Data, CRect* )

nMr,nSr, nMg,nMg, nMb,nMb = CStats:Mean( CImage, CRect* )

Data is a lua array (table) or a class object of type CImage, CArray, or CMatrix.
CRect is an optional CRect object that bounds of the computation for CImage and CMatrix objects.
nMean is the mean value
nSdev is the standard deviation.
nMr, nSr, etc. are the mean and standard deviation for each of the channels. If the CImage has 1 channel, 2 values are returned; if it has 3 channels, then 6 values are returned.
On success, the mean and standard deviation are returned for each channel.
On failure, returns nil, 0.

Remarks

Although this method returns two values, the mean and the standard deviation, you do not have to "catch" the standard deviation by assigning to a variable. For example,

m = S:Mean( T )	-- get the mean of table T.
m, sd = S:Mean( T )	-- gets the mean and standard deviation of table T.

If Mean is the last function in an expression, both values can be used implicitly, like

Printf( "%lg +/- %lg, Mean(t) ).

Example

Suppose a CImage I exists. The following script returns the mean and standard deviation of 600 pixels in the corner of the image:

S = CStats:new()	-- create a CStats object
m, sd = S:Mean( I, CRect:new( 1,20,1,30 )	-- computes values inside the CRect
Printf("mean=%lg +/- %lg\n", m, sd )	-- list the result
S:delete()	-- when done with S, delete it