LTPDA Toolbox™

contents

Version 2.9.4 LTPDA Toolbox Software

This table summarizes what's new in Version 2.9.4:

• Introduction
• Known Issues
• Main Changes
• Requirements
• New Methods
• Changes to methods
• MANTIS Issues Resolved
• Detailed changes since v2.8

Introduction

This version of LTPDA is 2.9.4. This document lists the changes since V2.8.

This version requires MATLAB 2012b or above.

Known Issues

None

Main Changes

This version of LTPDA contains many under-the-hood improvements, particularly in the area of speed, and also some changes to the underlying data structures for analysis objects.

This version also includes changes to the plotting tools to adapt to the changes to the graphics objects in recent MATLAB versions.

It is now possible to submit and retrieve MATLAB figures to/from an LTPDA repository, assuming the figures are created using LTPDA plotting tools (like iplot, ao/plot).

Spectral estimation methods now automatically drop the first N samples (defined by the window function) from the resulting estimates as these samples are biased. This can be deactivated in the plist. See help ao/psd for example.

LTPDA now supports data types other than double() in AOs, and some methods properly treat these data types.

Requirements

A number of changes made required us to change the minimum MATLAB installation requirements to

New Methods

• ao/mod An overload of the mod() function.
• ao/window A new method for windowing time-series data.
• ao/dtfe is a discrete transfer function estimation tool based on the DFT.
• ao/nyquistplot for making nyquist plots from fsdata AOs.
• ao/flscov performs iterative re-weighted least squares fitting.
• ao/summaryReport produces a tabular report of the input AO array. Currently only works for time-series AOs.
• ao/psdvfit vector fitting applied to spectra. It comes with a companion method ao/validateSpectrumMod for validating the fit.
• ao/union overloads the union operator for AOs.
• timespan(ao) a new constructor which creates a timespan object from a time-series AO.

Changes to methods

• ao/dft has been improved to allow windowing and scaling of the measurements.
• ao/interp has a new method called ‘auto’ which selects the best choice of interpolation based on the data type.
• ao/split now supports automatically splitting a time-series according to any significant gaps in the data.
• ao/resample now supports different data types. For non-double types, we do nearest neighbour interpolation.

MANTIS Issues Resolved

Release Notes

Version 2.9.3 LTPDA Toolbox Software

©LTP Team