Version history

Bug fixed in 'File: Export tree' procedure:

• Wrong edge identifications for bootstraps values
• Bootstraps values are now correctly exported !

Bug fixed in UPGMA/WPGMA tree construction procedure (Hierarchical clustering):

• Error with some dissimilarity files: Tree with "All edge length are null values !"
• Due to a bad accuracy coefficient estimation. Estimation procedure modified.

Bug fixed in 'File: Import sequence' procedure:

• Sensitive case Error in file verification

Modification in 'File: Export tree' procedure::

• If DARwin tree has bootstraps values
• Bootstraps values are now exported !

Bug fixed in 'File: Import tree' procedure::

• Wrong factor was applied in bootstrap values conversion

Syntax error correction fixed in:

• 'File: Export tree' procedure:
• 'Trees...: Max length tree' procedure:
• 'Trees...: Add a 2-degree node' procedure:
• 'Tools: Re-label trees for common identifiers' procedure:

Bugs fixed in 'Sampling for disequilibria: max length subtree' procedure:

• Bad taking in account of missing data values and error in case of negative missing date value
• Code adaptations for a better missing data mangement

Bugs fixed in 'Dissimilarity: Extreme values' procedure:

• Display error in case of external identifier selection, due to wrong array use
• Modifications to use the good array

Bugs fixed in Ward tree construction procedure (Hierarchical clustering):

• A compilation option in version 6.0.16, to optimize software execution, generate an infinite loop in the procedure...
• Code modification to make it compatible with the optimization option

Bugs fixed in some units selection procedures:

• Sampling for disequilibria:max length subtree
• Sampling for disequilibria:min SD subset

Bug fixed in 'Consensus tree and tree distances' procedure

• Error when great unit number in trees (overflow computing error)

Major optimizations in all units selection procedures:

• Include all interfaces with units or variables selection
• Adding real numeric sort algorithm with aumoatic detection of identifier type
• Optimization of time in sort procedures
• Optimization of time in select/unselect and display procedures

Bug fixed in 'Tree contruction Hierachical Clustering'

Methods 'UPGMA' and 'WPGMA' procedure:

• Error with some dissimilarity file (overflow computing error)

Bug fixed in tree contruction 'Neighbor Joining under topological constraints' procedure:

• Error with some dissimilarity file (overflow computing error)

Minor bugs fixed in dissimilarity computing:

• Error with big data matrix (more than 30,000 variables)

Improvements in dissimilarity calculation from single data:

• Multi-Threading added in missing data exploration procedure

Improvement in file selection (all procedures):

• Adding error handling on file already open by another procedure
(return to top of page)

Minor bug fixed in external identifiers reading procedure:

• Error when Unit number greater than 32.000

Minor bug fixed in external identifiers writing procedure:

• Error when Unit number greater than 32.000

Improvement in application closing procedure:

• Adding error handling on application parameters recording

Bug fixed in Pooling SSR alleles procedure:

• Error in Unit selection window while selecting Identifiers file

Bug fixed in Import Data procedure:

• Error handling when the file to import is open in other software (like Microsoft Excel)

Unit selection window

called by dissimilarity calculation, single data correlations, Export for phase, Pooling SSR alleles :

• Major improvement of design speed
• Minor bug fixed

Application compilation option modified:

• Fixes a dead end in 'Tree construction-Hierarchical Clustering-Ward' procedure
• Enables error control in some procedures

Bug fixed in 'Tree construction-Hierarchical Clustering-Ward':

• Error in some dissimilarity configurations

Minor bug fixed in Import data matrix procedure:

• Input file validation malfunctioning

Bug fixed in 'Calculate dissimilarity from single data' for counts index 'Chi 2':

• Error in dissimilarity final computing: all data matrix are set to zero

Bug fixed in 'Calculate dissimilarity from single data' (for all indexes):

• Error while setting ProgressBar initial values with particular variables number or computer's core number

Minor bug fixed in Import data matrix procedure:

• Error handling when the file open in other procedure

Minor improvements in 'Calculate dissimilarity...' (for all data type) bootstrap computing:

• In case of aborted procedure the window does not close and stays enabled to modify options and retry computing

Minor improvements in maximized windows handling:

• Fixed window opening reduces automatically maximized window if needed

Bug fixed in 'Tree construction-Hierarchical Clustering-Ward':

• Error in edges lengths

Minor bug fixed in 'Factorial analysis-Graphical display':

• Axis number was not correctly refreshed

Minor bug fixed in 'Tree-draw' and 'Factorial analysis-Graphical display':

• Graphics don't work with font size less or equal to 2 points

Bug fixed in 'Calculate dissimilarity …' (for all data type):

• Error while selecting bootstraps number from 1000 onwards (only on English windows configuration)

Minor improvements on Tree Drawing:

• Margins was not correctly taken in account in the print procedure
• Cancel button on form closing had no effect

Minor improvements on Factorial analysis Drawing:

• Margins was not correctly taken in account in the print procedure

New procedure in 'File' menu:

• 'Export dissimilarity as column' text file

Bug fixed in 'Calculate dissimilarity from allelic data':

• Controls re-activation in case of procedure aborting
(return to top of page)

Important bug fixed in multi-threading procedures:

• computing instabilty in particular cases

Improvements on 'Units selection':

• store .DON file name and selected variable for reuse

Bugs fixed in 'Calculate dissimilarity from single data':

• 'Record selected subset...' option
• computing of dissimilarity indices for 'modalities' option

Improvements on application management:

• Check for the availability of a new version on DARwin website

Bug fixed in Import data matrix procedure:

• Ploidy recording in case of allelic data

Bug fixed in Import dissimilarity procedure:

• Importation from Phylip format

Minor improvements in AFDT and Tree representation:

• Force redraw which did not occur when maximizing window on some computers

This new version has been completely rewritten under a more recent system providing support for addressing the multiple cores of the multicore processors. All the procedures have been parallelized in order to exploit multi-threading, with as consequence a time execution decrease proportional to the number of cores.

Development is now written in Visual Basic with Microsoft Visual Studio 2010. The distributed software runs on Microsoft .Net Framework 4.0 under 32 and 64 bits platforms.

Upgrade of software data capability for input VAR files and other procedures:

• Over 32.000 units
• Over 32.000 variables

Minor improvements in Disequilibrium procedures:

• Adding memory capacity overflow tests

Minor improvements in Disequilibrium procedures:

• Adding memory capacity overflow tests

Major improvements in:

• EMF tree export procedure
• EMF AFTD export procedure

Minor improvements in:

• UnWeighted Neighbor-Joining tree construction procedure
• Extreme dissimilarity values procedure

File menu:

• New procedure: Import data matrix

Dissimilarity menu:

• New procedure: Extreme values

Reticulations:

• add reticulations to a tree to improve the LS fit between graph distances and observed dissimilarities.

Factorial Analysys - Analysis:

• Bug correction in unit list: unit #1 not taken into account
• only in V 5.0.152 and V 5.0.153

Tools - single data correlations:

• New option: optional correlation matrix record as dissimilarity file

Dissimilaty - Transformations:

• New option: "1 - D" (Similarity to Dissimilarity transformation or conversely)

• Bug correction in prelimirary calculation: error #6 in tree draw due to all edge lengths set to null value or edge lengths set to 1)

in Tree construction menu:

• Neighbor Joining
• Neighbor Joining under topological constraints
• Influencial unit detection

• Bug correction in prelimirary calculation (error #6 in tree draw due to all edge lengths set to null value)

in Tree construction menu:

• Neighbor Joining
• Neighbor Joining under topological constraints
• Influencial unit detection

Neighbor Joining tree construction procedures:

• accuracy improvement in case of low dissimilarities values

in Tree construction menu:

• Neighbor Joining
• Neighbor Joining under topological constraints
• Influencial unit detection

• local optimizations

• physical zoom to freely magnify part of the graph

• option to mark alleles as missing

• Pooling SSR alleles

• Ploidy=1 (haploid) is allowed

• bug fixed: edge length estimation in case of the dissimilarity values at the same precision

- Called by:

• Tree construction
• Neighbor Joining under topological constraints

• Bug fixed: Generated tree file was not a valid tree in case of line feed in the input file

• Bug fixed: elements out of the frame

Called by:

• Tree…
• Draw

• SD computing modules:
• Sampling for disequilibria: max length subtree
• Sampling for disequilibria: min SD subset
• Optimization: global reduction of computing time in case of large number of units and locus

• Bug fixed: memory overflow error for trees on very large number of units

called by:

• Tree…
• Tree distance
• Fit criterion
• Edge length global estimation
• Max length subtree
• Diseq. Menu
• Sampling for disequilibria: max length subtree
(return to top of page)