Skip to content

Releases: kegge13/BistroMath

single non-recurrent memory leak resolved

29 May 13:40
7edcabc
Compare
Choose a tag to compare

On termination the primary analysis engine was not freed from memory. Therefore a lot of memory leaks were reported.

Wedge alignment bug and refactoring

20 Feb 21:04
Compare
Choose a tag to compare

Minor bug fix: Even with Wedge detection on, alignment could lead to a wrong center position for very wide wedged fields and resulting in unreliable alignment for the reference. This could be corrected with a reload of the profile. Now the edge found through the derivative is used when for the derivative is concluded that a wedged profile is found.

Slightly faster analysis by avoiding recalculation of cpu-intensive results if already available.

A CPU load indicator and a Field Type have been added to the statusbar.

The functionality of the Processing option Mirrored measurement as Reference is extended to all profiles in the history list.

Gamma analysis step size bugfix

27 Jan 10:14
Compare
Choose a tag to compare

The conversion from the UI steps per mm to the calculation's stepsize in cm a multiplication of 10 was applied instead of division. This error was existent since the migration from Delhi to Lazarus.
See also the release information on the 4.30.1 version below.

Focus Position

22 Jan 14:19
Compare
Choose a tag to compare

The Focus position can be analysed from a series of measurements. In this subversion the list of supported devices for structured multiple data sets is accessible in the user interface.
FocusPositionTab

A sigmoid analysis is used to find the center of each profile. This is much more stable for 2D arrays. Although the sigmoid point deviates from the classic 50% position, the average of left and right edge will give the correct result. In the graph below only the right border values are shown.
shift_analysis_TP_2010

sigmoid slope normalisation

11 Jul 10:52
7fd6c9d
Compare
Choose a tag to compare

After discussions in the NCS-33 subcommittee it is seen as most appropriate for reporting purposes to normalise the slope of a sigmoid curve to its local value instead of the global normalisation value.

statistics bugfix

20 Jun 08:09
efe1012
Compare
Choose a tag to compare

Due to the limitations of the 64 bits floating point arithmatic a small series of identical samples might give a non-zero, negative variance in the statistics module. This could be triggered in Gamma analysis.
When in generic data positions are interpreted as mm, a warning is issued.

update fof LCL 2.2.2 / FPC 3.2.2

23 May 12:48
152795e
Compare
Choose a tag to compare

Tiny changes in logging output.
Clarified text in user interface

Integrate function bug fix

21 Apr 21:37
d1ed4ae
Compare
Choose a tag to compare

Bug fix for Integrate function
Accidently dropped first data point for generic (two columns of numbers) data.

FFF support in selected files

15 Apr 12:01
7aca3fa
Compare
Choose a tag to compare

Set FFF tag manually for any scan.
Find FFF tag in OmniPro v6/7/8, RFA300, SNC clipboard, SNC .snctxt and PTW.

allow pdd normalisation at zero depth

21 Feb 12:16
278dc38
Compare
Choose a tag to compare

In VerticalScans is assumed that the norm position is >0. This is the case for an open water interface. At angle 90/270 this might not be valid.