January 23, 2015: Version 1.32: Switched to autobuild tools for
building vartools.

January 20, 2015: Version 1.31: The following new commands have been
added: -BlsFixDurTc, -copylc, -nonlinfit, -resample, and
-WWZ. Significantly revised -addnoise command allowing simulation of
time-correlated noise via a Gaussian process model. Modified routines
for loading user-developed libraries (initialization function now
requires more input values). Auto sort light curves on input. Added
"wmedian" and "wpct" statistics to -stats command. Added "niter" and
"median" options to -clip. Added "format" option to -linfit.

September 11, 2014: fixed bug causing buffer overflow or seg-fault
when outputting the model function in the -Killharm command. Thanks to
T. Bovaird for the bug report.

July 16, 2014: corrected bugs related to compiling without libraries,
and to compiling on windows.

June 17, 2014: fixed bug in vartools_functionpointers.h preventing
compilation with PARALLEL option turned off.

April 22, 2014: Version 1.3: Major update. The following new commands
have been added: -changevariable, -expr, -if -else -elif -fi, -linfit,
-restricttimes, -stats. The following new options have been added:
-inputlcformat, -inlistvars, -showinputlistformat, -showinputlcformat,
-log-command-line, -functionlist, -L and -F. The following new
extension commands have been added: -fastchi2, -jktebop, -macula,
-magadd. Changed the "spec" keyword to "list" for all commands, which
is hopefully a bit more intuitive. Changed the syntax for -o to allow
greater control over the formatting of the output. Changed -aov_harm
to correct for a singularity which can appear when applying the method
to evenly sampled data (thanks to A. Schwarzenberg-Czerny for pointing
this out). Changed -binlc and -clip to handle all light curve related
vectors, not just t, mag and err. Allow reading in rows of arbitrary
length from ascii files. Significant new features include: the ability
to handle analytic expressions and to assign variable names to input
vectors and scalars; the ability to read-in many columns from a light
curve as vectors and to change the vectors which commands will operate
on with the -changevariable command; support for user developed
extensions (including adding your own commands to vartools, or adding
new functions to be used with -expr, -linfit or other commands which
handle analytic expressions).

July 18, 2013: added Mac installation notes. Thanks to J. Pepper and
his research team for providing this.

May 15, 2013: updated help for the -Jstet command to note a difference
between the J statistic computed by vartools and that defined in
Stetson's paper. Thanks to L. Macri for reporting this.

Apr 23, 2012: Version 1.202: Fixed bugs in -converttime related to
location of observer and handling of input-ppm, also corrected bug in
-converttime causing loss of timing precision from accumulated
rounding errors. Time conversions should now by precise to ~0.1
millisecond near JD2000.0. Reworked the internal method for parsing
light curve data, if bugs are encountered the "-D _USE_PARSELC" term
in the makefile can be deleted to return to the old method. This
change should not affect the behavior of the program, but should make
future developments (e.g. support for multi-filter or multi-aperture
light curves) easier. Changed -readformat to allow reading in light
curves that are missing one or more of the JD, mag, or sig
columns. Fixed memory leak in -getampthresh. A lot of code has been
added related to supporting user-developed extensions to vartools; at
present this feature is not fully functional and by default is not
compiled into the program.

Feb 8, 2012: Fixed bug in -decorr causing the LCColumn_?_coeff_err_?_?
to not be output correctly. Thanks to D. Flateau for the report.

Feb 3, 2012: Version 1.201: Fixed bug in -BLSFixPer causing incorrect
periods to be printed to ascii output table when using "list". Thanks
to J. Pepper for the bug report.

Feb 1, 2012: Fixed seg-fault when using -inputlistformat with -SYSREM,
also corrected a bug in the output table when combining -header and
-redirectstats. Thanks to D. Flateau for the bug report.

Nov 10, 2011: Updated descriptions of some commands on the website.

Oct 21, 2011: Version 1.2. Significant changes including: added
"-addnoise" and "-converttime" commands, and "-example",
"-inputlistformat" and "-parallel" options. Significant internal
revisions to the code to allow for parallel processing of light curves
(affecting many commands which had previously used global or static
variables; there may be bugs introduced as a result, please let me
know if you encounter unexpected behavior). Added options to most
commands to specify the column number for parameters taken from the
input list file. Added "inpututc" option to "-readformat". Changed
"-help" to only provide a brief synopsis of the program, "-help all"
gives the previous behavior of "-help". Changed -MandelAgolTransit and
-SoftenedTransit to internally vary the initial and final transit
epochs rather than the period and epoch, also improved the initial
parameter estimates based on -BLS. Added "dilute" option to
-Injecttransit to simulate diluted transits.

Apr 6, 2011: Version 1.158. Added "fittrap" option to BLS and
BLSFixPer which fits a trapezoid transit at each peak period. Added
"nobinnedrms" option to BLS. This option speeds up the running of BLS,
but results in a BLS_SN statistic that is suppressed. Use this option
if not using BLS_SN to select transits. Added ophcurve and ojdcurve
options to BLS. Fixed a bug in BLS which caused period "-1" peaks to
be listed before the correct peaks in cases where fewer than Npeaks
periods were found. Added "-oneline" option which provides easier to
read output when processing a single light curve.

Apr 5, 2011: Version 1.157. Fixed bug in BLS causing integer overflow
for large frequency number searches. Thanks to B. Sipöcz for pointing
out the bug, and the fix.

Mar 1, 2011: Version 1.156. Fixed bug in BLS causing memory
leak. Thanks to R. Siverd for debugging this.

Feb 25, 2011: Version 1.155. Fixed bug in BLS causing the wrong
transit time center to be output. Thanks to T. Beatty for reporting
the bug.

Feb 14, 2011: Version 1.154. Minor change to inputoutput.c to correct
unpredictable behavior when reading fits files, thanks to J. Rasor for
reporting the bug. Fix bug relating to the file names of bls
periodograms and added version number to the help and usage functions
(thanks to R. Siverd for the report/suggestion).

Feb 10, 2011: Version 1.153. Added the ability to read in binary fits
table files. Renamed old -fluxtomag command as -difffluxtomag and
added a new -fluxtomag command. Added option to use amplitude spectra
with -dftclean. Added replace option to -medianfilter command. Some
improvements to the -Phase command. Fixed a bug in -starspot which
prevented the model from being output when no fitting was
done. Modified output of -BLS and -BLSFixPer commands to include the
epoch of transit center for the first transit after the start of the
light curve.

Aug 13, 2009: Version 1.152. Added the -microlens command, changed the
DFT routine in -dftclean from the brute force to the FDFT algorithm,
fixed a bug in the calculation of S/N for the -LS command, added GNU
license.

May 19, 2009: Version 1.151. Added the -findblends command (thanks to
Kris Stanek for the suggestion).

Mar 3, 2009: New Version 1.15. Changes include: 1. Added whiten, and
clip options to aov, aov_harm and LS commands. Also added SNR output
to these commands, and the option to determine the SNR at a fixed
period for these commands. Note that the default behavior now for -aov
is to give the AoV statistic and the SNR of the AoV statistics,
previously the behavior was to give the SNR of the log AoV statistic
which tended to suppress the detection significance for large
amplitude variables. To get the previous behavior use the "uselog"
option for this command. 2. Added the -Injecttransit command. 3. Fixed
a bug in the aov peak finder that could cause it to miss the peak in
some cases. 4. Fixed a bug in the -MandelAgolTransit algorithm that
caused the transit to not appear at the correct phase for some values
of omega. Changed -MandelAgolTransit to input/output omega in
degrees. 5. Fixed an incompatibility between the -clip command and
several other commands. 6. Added the -savelc and -restorelc
commands. 7. Changed -autocorrelation to use the Discrete Correlation
Function (binning is done on the ACF rather than on the lc, this also
yields error estimates for the ACF). 8. Added the ability to reference
arbitrary previously computed statistics by output column name or
number. This option has not yet been fully extended to all commands
that can use results from previous calculations. 9. Added the option
of reading in signals from a file to the -GetLSAmpThresh
command. 10. Added the option of simultaneously fitting a fourier
series to a light curve with TFA to the -TFA_SR command.

Dec 11, 2008: Changed -Killharm to allow different output formats,
fixed a bug with the phaserel and amprel options to -Injectharm, added
an example on injecting and recovering a simulated RR Lyrae signal.

Dec 10, 2008: Substantial changes. Added several new commands:
-autocorrelation, -changeerror, -dftclean, -Injectharm, and
-TFA_SR. Added the following new options: -matchstringid, -quiet,
-randseed, -numbercolumns, -listcommands. Changed the -help option to
allow displaying help for individual commands. Modified Killharm to
allow "injectharm" and "fix" options for inputting periods, also added
the "fitonly" option and changed the output format/column headings for
this command. Modified TFA to allow specification of readformat of
trend lcs. Added fitRV options to -MandelAgolTransit, this hasn't been
fully debugged yet, so I suggest not using this feature for the
moment.

Sep 25, 2008: Changed the "Mean_Mag" output for the -Killharm command
to be the fitted mean rather than the statistical mean, this is now in
line with the way this was presented in Example 2 above.

Sep 22, 2008: Added -nobuffer option (thanks to Rob Siverd).

Aug 18, 2008: Fixed seg. fault when reading long input lists (thanks
to Josh Pepper).

July 31, 2008: Added -headeronly option.

July 2, 2008: Modified to allow input from stdin.

May 27, 2008: Modified the -BLS and -BLSFixPer commands to output the
number of points in transit, the number of transits, the number of
points before and after transit, the red and white noise of the signal
after removing the bls model and the signal to pink noise ratio of the
transit.

May 25, 2008: Fixed seg. fault in initialize_tfa routine. Added
examples. (thanks to Rob Siverd)

Mar 26, 2008: Fixed memory leak in -aov command. (thanks to David Nataf)

Jan 16, 2008: Fixed isinf portability problem. (thanks to Alceste Bonanos)

Nov 26, 2007: Added -SYSREM and -binlc commands and the -jdtol option.

Nov 19, 2007: Added -TFA and -fluxtomag commands.

Nov 15, 2007: Added -MandelAgolTransit, -GetLSAmpThresh, and -aov_harm
commands.
