10/11/2021: v0.6.1-9001

  o reduce memory footprint of
	procSnps: working on rcmat directly instead of creating out
	counts2logROR: working directly on out (input nam now instead of mat)
  o added rePreProcSample to help reanalysis starting from procSample output

10/08/2021: v0.6.1-9001

  o Register native API.
  o Fix partial argument and element names.

10/30/2019: v0.6.1

  o fixed the emcncf "missing value where TRUE/FALSE needed" error
  o modified jointsegsummary to be faster (using indices instead of tapply)
  o procSample modified to get the cumulative segment indicator from seg

12/13/2018: v0.6.0

  o Removed emcncf2 and marked as defunct
  o Changed the order of tcn-lcn lines being drawn to address (0,0) masking
  o Added udef option for gbuild to enable analyzing other genomes (say
      dog) with user supplied GC percentage data

02/28/2018: v0.5.14

  o Deprecating emcncf2 to prepare for removal in the next version.

12/08/2017: v0.5.13

  o Added the diagnostic spider plot for logR-logOR summary of segments.
  o function readSnpMatrixDT to read the pileup count file using
      data.table package added to extRfns directory (Thx Dario Beraldi).

10/18/2017: v0.5.12

  o documentation to explain the need to set the seed and NA in cncf.

03/31/2017: v0.5.11

  o added vertical lines to indicate chromosome boundaries

03/28/2017: v0.5.10

  o fixed a bug in plotSample. EM fit doesn't have naive tcn, lcn and cf.

03/24/2017: v0.5.9

  o added stomach example to tests for code coverage

03/23/2017: v0.5.8

  o nhet threshold checks in emcncf and emcncf2

03/21/2017: v0.5.7

  o added details of the columns of the segment summary data frame

11/11/2016: v0.5.6

  o fixed bug in readSnpMatrix which removed all loci with any del or
    err read. New default is to keep all.
  o fixed bug in cbs2rank.f code that was greedy in search for max T.
  o changes cbs2rank.f code and R functions (segsnps etc. that call it)
    to specify delta - min change in E(logR).
  o preProcSample has a new option delta now to implement
  o readSnpMatrix can read the old perl pileup files now
  o fixed integer overflow bug in fit.cpt.tree (11/18/2016)

10/28/2016: v0.5.5

  o fixed bugs in emcncf and emcncf2 (affected cases with no CNA)

10/18/2016: v0.5.4

  o added colClasses to read.csv in readSnpMatrix (Henrik's comment)
  o changed stomach.csv.gz to have Ref & Alt columns (all "N")

09/22/2016: v0.5.3

  o added hg38 as a genome build option (needs pctGCdata >= 0.2.0)
  o changed the code to allow for chromosomes with no snps

08/10/2016: v0.5.2

  o fixed the bug in counts2logROR that did not pass gbuild
  o included snp-pileup code in inst/extcode directory. This
    can be compiled (needs htslib) to create a stand-alone
    read counts snpmatrix file.

06/15/2016: v0.5.1

  o added ndepthmax to allow for high coverage samples
  o fixed ChrX=23 coding in cncf & emcncf to handle mouse (chrX=20)

06/03/2016: v0.5.0

  o this is a major revision with changes to the interface.
  o versions through 0.3.30 were released with bug fixes
  o reading of count matrix is done in a separate function
  o plotSample is versatile and plotSampleCNCF removed

01/11/2016: v0.3.20

  o findDiploidLogR fixed to account for 2+2 states at 2 cnlr levels

01/08/2016: v0.3.19

  o some adjustments to cncf so that max fitted cf is 99%

01/08/2016: v0.3.18

  o fixes to emcncf when estimated naive cf is small

01/08/2016: v0.3.17

  o segments should have 1% of hest for mafR clustering
  o check for since chrX in cncf

01/07/2016: v0.3.16

  o trimmed fit for emcncf

01/07/2016: v0.3.15

  o add chr X check

01/07/2016: v0.3.14

  o further refinement of emcncf

12/22/2015: v0.3.13

  o refinement of emcncf; update plotSampleCNCF.Rd

12/09/2015: v0.3.12

  o purity optimization edit

12/03/2015: v0.3.11

  o purity optimization take2; update DESCRIPTION

12/02/2015: v0.3.10

  o fixed purity optimization

11/17/2015: v0.3.9

  o Bug in winsorized mafR code (median(abs(valor)) instead median(valor))

09/09/2015: v0.3.8

  o Modified balanced seg selection restrict to  mafR < 0.1
  o Winsorize maf calculation to dampen extreme valor values.
  o Added unmatched normal code.

08/14/2015: v0.3.7

  o Fixed negative copy numbers in emcncf

08/04/2015: v0.3.6

  o Added dipLogR argument to procSample to help reprocess with lower cval.

07/30/2015: v0.3.5

  o Changed fitcncf to use more stringent deviation measure.

07/24/2015: v0.3.4

  o Fixed bugs in emcncf (Ronglai)

07/24/2015: v0.3.3

  o default min.nhet set to 15. changes to findDiploidLogR and clustersegs.

07/22/2015: v0.3.2

  o Fixed bugs in emcncf (Ronglai)

07/21/2015: v0.3.1

  o Added checks to findDiploidLogR, segsnps and fit.cpt.tree. Fixed
	emcncf bug. Set default cval to be 50 in procSample.
