Below is a brief descriptions of the spectral analysis and statistical functions provided in the TeraScan system.
The spectral analysis functions are mainly composed of:
lagspect - computes the power spectrum from the correlation function. The correlation can be a one-dimensional auto- or cross-correlation. In general, the input is expected to be output from xcorrel.
randata - generates 1- or 2-dimensional variables of random numbers. This function is provided for testing purposes or for use in statistical testing.
spect - computes the power/phase spectra of a variable with respect to one dimension. The data can be multi-dimensional, however only one dimension is transformed at a time (see spect2d). Options are provided for band-pass filtering the transform and then inverse transforming to obtain a filtered version of the input.
spect2d - computes the power/phase spectra of two-dimensional variables. Options are provided for band-pass filtering the 2D transform and then inverse transforming to obtain a filtered version of the input.
xcorrel - computes auto- and cross-correlations for variables. The data can be multi-dimensional, however the correlation can only be processed for one dimension at a time. Note: if data contain bad values, power spectra can often be obtained by transforming (via lagspect) the auto- or cross-correlation functions.
xspect - computes the power/phase cross-spectra of two variables with respect to one dimension. The data can be multi-dimensional, however only one dimension is transformed at a time.
The statistical analysis functions are mainly composed of:
binavg - computes the average relationship between two variables, i.e. how does variable A relate to variable B. Variable A is bin averaged according to the corresponding values of variable B. The number used in the bin average can be reported in addition to the standard deviation.
bvhist - computes the bivariate histogram for two variables.
correlate - computes a correlation matrix between a set of variables in a dataset.
dimavg - average variables over a specified dimension.
emath - general purpose function for performing mathematical operations between variables.
fakedata - used to generate test and artificial data.
grad - computes the gradient of a variable with respect to a specified dimension. Gradients can be based on dimension values or actual Earth distances.
smear - performs data smoothing of a variable.
stats - computes simple statistics for variables such as mean, min, max, and standard deviation.
eof, binavg, bvhist, correlate, dimavg, emath, fakedata, grad, intro, lagspect, randata, smear, spect, spectd2d, stats, xcorrel, xspect.
Last Update: $Date: 2000/11/20 19:12:06 $