Changes to version 0.0-38

	o added check for very short trips that break the windowing in position.logp

	o renamed CHANGES to NEWS

Changes to version 0.0-37

	o reverted to Suggests for locfit, since that package has no namespace

Changes to version 0.0-36

	o added a package NAMESPACE

	o reversed order of CHANGES file entries

	o fixed mis-order in warning message use of CRS/CRSargs in
	satellite.model and solar.model


Changes to version 0.0-35
	
	o Sanity test for NAs and class in the data for mkLookup. 
		
Changes to version 0.0-34
	
	o Added functions get.sst, sst.interp and sst.interp1. 
	
Changes to version 0.0-33


	o added functions mkMaskObject and "set.mask<-"

	o added winoffset argument to position.logp

	o Modified the algorithm used by position.logp
	
Changes to version 0.0-30


	
	o added time vector to output of solar.model to make keeping track of the values simpler
	
	o Fixed bug in position.logp with incorrect check for is.null(start) 
	
Changes to version 0.0-30
	
	o Fixed stray boilerplate "pkg" references in .Rd files. 
	
Changes to version 0.0-29
	
	o New build version to get around problems submitting to CRAN. 

Changes to version 0.0-28

	Thanks to Nick.Ellis@csiro.au for the report.  2009-05-29
        ported from Montenbruck and Pfleger (1994).  
        between the values from elevation() and the astro() functions
        o Fixed an embarassing error in POLAR that caused a disparity
        
        o Added transformation capacity to satellite.model
        
        o Added proj.string to satellite.model
        
        warning message about logical to double type conversion.
        o Fixed broken checks in solar.model that resulted in worrying
        
        match solar.model
        satellite.model with behav.dist, behav.mean and behav.sd to
        o Replaced speed.dist, speed.mean and speed.sd arguments in
        
        replaced by pmax.int, pmin.int
        o Removed fast.pmax and fast.pmin from solar/satellite.model,
        
	probability control (new prob arg)
	o Updated position.logp to more sensibly allow quantile
	
Changes to version 0.0-27
	
	o Fixed a bad mistake from the last build in solar.model
	
Changes to version 0.0-26
	
	
	David Pinaud)
	o Fixed missing transform function from solar.model (thanks to
	
Changes to version 0.0-25
	
	
	
	solar
	
	elevation
	
	julday
	julcent
	
	---------
	SOLAR
	
	solar.model
	satellite.model
	
	norm.proposal
	mvnorm.proposal
	bmvnorm.proposal
	
	metropolis0
	metropolis
	
			## k.prior DELETED
	------------------
	GEO.METROPOLIS
	
			##reynolds DELETED
			##podaac   DELETED
	mkLookup 
	
	mkTempMask
	mkSmall
	mkDepthMask
	get.mask
	
			## fetch.topo DELETED
	
	bits<-
	bits
	--------
	MASKS
	
	
	picksegs
	pick
	
	mkCalibration
	
	show.segment
	light.quantile
	position.logp
	initialize.x
	
	
	----------------
	MISCELLANEOUS
	
	old.mkLookup
	old.metropolis
	old.find.init
	old.dist.gc
	mkNLPosterior
	mkElevationSeg
	------------------
	OLD GEO.METROPOLIS
	
	POLAR
	MJD
	mini.sun
	lunar
	LMST
	FRAC
	EQUHOR
	astro
	-----------
	MINI ASTRO
	
	
		
	R files
	shows the grouping of functions in doc, sometimes also in the
	o restructured package with documentation, the following list

Changes to version 0.0-17

	
	
	distance functions appropriately - km units are assumed
	o solar.model also accepts a proj4 string, and creates
		
	behav.sigma
	model - the arguments have become behav, behav.mean and
	o solar.model now has support for "distance" as well as speed
	  
	  elevation "ekstrom" arguments
	  outside a given range, this is specified either by light or
	o solar.model can now put a larger variance on light levels
	
	o show.segment now has "..." to pass in graphics options
	
	class for imaging the array
	o position.logp has some changes in that I've added an S3
	
	o metropolis now reports on progress

	needs some fixing in terms of naming etc.)
	intersections of large logp values (not the best function yet,
	o init.x.diag - added function to initialize off running
	
	o mkCalibration now allows a single segment

	o Removed model.utils and put it all in solar.model
Changes to version 0.0.10

	o I'm testing automatic Subversion updating to R repository
Changes to version 0.0.9

 
	o fetch.topo now uses rgdal to read an ESRI FLT surface
	
	default is they are not passed in, but initialize occurs after
	o Changed operation of start.x and start.z in solar.model -
	
Changes to version 0.0-8

	o Added function findSegments
	
Changes to version 0.0-7


	logp.behavioural
	o fixed some bugs (not very well) in mkLookup and
	
Changes to version 0.0-6

	of the time
	o changed defaults in model.utils to avoid require(rgdal) most
	
Changes to version 0.0-5

	o incorporated new functions for diagnostics
	
Changes to version 0.0-4

Changes to version 0.0-34
		
	o Sanity test for NAs and class in the data for mkLookup. 
	
Changes to version 0.0-33
	
	o Added functions get.sst, sst.interp and sst.interp1. 
	
Changes to version 0.0-30
	
	o Modified the algorithm used by position.logp

	o added winoffset argument to position.logp

	o added functions mkMaskObject and "set.mask<-"


Changes to version 0.0-30
	
	o Fixed bug in position.logp with incorrect check for is.null(start) 
	
	o added time vector to output of solar.model to make keeping
	track of the values simpler


Changes to version 0.0-29
	
	o Fixed stray boilerplate "pkg" references in .Rd files. 
	
Changes to version 0.0-28

	o New build version to get around problems submitting to CRAN. 
	
Changes to version 0.0-27
	
	o Updated position.logp to more sensibly allow quantile
	probability control (new prob arg)
        
        o Removed fast.pmax and fast.pmin from solar/satellite.model,
        replaced by pmax.int, pmin.int
        
        o Replaced speed.dist, speed.mean and speed.sd arguments in
        satellite.model with behav.dist, behav.mean and behav.sd to
        match solar.model
        
        o Fixed broken checks in solar.model that resulted in worrying
        warning message about logical to double type conversion.
        
        o Added proj.string to satellite.model
        
        o Added transformation capacity to satellite.model
        
        o Fixed an embarassing error in POLAR that caused a disparity
        between the values from elevation() and the astro() functions
        ported from Montenbruck and Pfleger (1994).  
	Thanks to Nick.Ellis@csiro.au for the report.  2009-05-29

Changes to version 0.0-26
	
	o Fixed a bad mistake from the last build in solar.model
	
Changes to version 0.0-25
	
	o Fixed missing transform function from solar.model (thanks to
	David Pinaud)
	
	
Changes to version 0.0-17

	o restructured package with documentation, the following list
	shows the grouping of functions in doc, sometimes also in the
	R files
		
	
	
	MINI ASTRO
	-----------
	astro
	EQUHOR
	FRAC
	LMST
	lunar
	mini.sun
	MJD
	POLAR
	
	OLD GEO.METROPOLIS
	------------------
	mkElevationSeg
	mkNLPosterior
	old.dist.gc
	old.find.init
	old.metropolis
	old.mkLookup
	
	MISCELLANEOUS
	----------------
	
	
	initialize.x
	position.logp
	light.quantile
	show.segment
	
	mkCalibration
	
	pick
	picksegs
	
	
	MASKS
	--------
	bits
	bits<-
	
			## fetch.topo DELETED
	
	get.mask
	mkDepthMask
	mkSmall
	mkTempMask
	
	mkLookup 
			##podaac   DELETED
			##reynolds DELETED
	
	GEO.METROPOLIS
	------------------
			## k.prior DELETED
	
	metropolis
	metropolis0
	
	bmvnorm.proposal
	mvnorm.proposal
	norm.proposal
	
	satellite.model
	solar.model
	
	SOLAR
	---------
	
	julcent
	julday
	
	elevation
	
	solar
	
	
	
Changes to version 0.0.10
	o Removed model.utils and put it all in solar.model

	o mkCalibration now allows a single segment
	
	o init.x.diag - added function to initialize off running
	intersections of large logp values (not the best function yet,
	needs some fixing in terms of naming etc.)

	o metropolis now reports on progress
	
	o position.logp has some changes in that I've added an S3
	class for imaging the array
	
	o show.segment now has "..." to pass in graphics options
	
	o solar.model can now put a larger variance on light levels
	  outside a given range, this is specified either by light or
	  elevation "ekstrom" arguments
	  
	o solar.model now has support for "distance" as well as speed
	model - the arguments have become behav, behav.mean and
	behav.sigma
		
	o solar.model also accepts a proj4 string, and creates
	distance functions appropriately - km units are assumed
	
	

Changes to version 0.0.9
	o I'm testing automatic Subversion updating to R repository

Changes to version 0.0-8
	
	o Changed operation of start.x and start.z in solar.model -
	default is they are not passed in, but initialize occurs after
	
	o fetch.topo now uses rgdal to read an ESRI FLT surface
 

Changes to version 0.0-7
	
	o Added function findSegments

Changes to version 0.0-6
	
	o fixed some bugs (not very well) in mkLookup and
	logp.behavioural


Changes to version 0.0-5
	
	o changed defaults in model.utils to avoid require(rgdal) most
	of the time

Changes to version 0.0-4
	
	o incorporated new functions for diagnostics

Changes to version 0.0-3 

	o Fixed a bug in reynolds, it needs some stray integers stripped out
	(old Fortran something) - comments in the code show how to
	replicate the test output from the site.  o Added capability
	for gzipped files to reynolds, as well as "data" option to
	return sst, ice, err, land or water.
	
	o Added k0 attenuation offset to find.init - seems to perform
	better.

	o Fixed a bug! in mkTempMask - it was using only the
	initialized Reynolds dataset.


Changes to Version 0.0-2
o Recreated package completely

Changes to version 0.0-1
o Created package with all latest code

