The following bugs were fixed in PSTAT 6.0.1
============================================

*  The number of fitting components in the leVel/Setup table now 
   accepts the correct choice of 1-6 fitting components.  Previously only 
   0-5 could be entered.

*  Analyze/Export/Plot in binned-only storage mode now works correctly.
   Previously this operation may have caused a program crash with some files.

*  Data limits are now correctly reported in Analyze/Fit/Statistics.  
   Previously the upper data limit was one bin width too low.

*  Re-binning now works correctly if a very large number of look-ahead 
   bins is selected.  Previously this could have caused a program crash.

*  The command line file is now automatically processed.  It was 
   previously ignored.

*  The time scale field in Analyze/Type/Ideal form is now correctly 
   formatted.  There was no formatting previously.

*  After viewing help in the Export/Plot/View menu, the program 
   execution now returns to the menu.  Previously this caused a program 
   crash.

*  pSTAT now handles empty histograms in logarithmic binning mode 
   correctly.  Empty histograms in this mode previously caused a 
   program crash.

*  The "Total time for this level" in Analyze/Information now correctly 
   reads "(unavailable)" for amplitude data, rather than displaying a 
   meaningless value as previously.

*  The arithmetic means for dwell-time and amplitude data are now 
   reported correctly in the Analyze/Information screen.

*  The hardcopy plot is no longer clipped with A4 paper size and 100% 
   scaling.

*  A dotted line rather than a solid line is now plotted on the hardcopy 
   for an intermediate fit.

*  The full fitting curve is now drawn if the X-axis is logarithmic.  Only 
   a portion was previously drawn.

*  The X-axis is now drawn correctly on the Y-axis zero (if optionally 
   selected) in logarithmic binning and in the scatter plot.

*  The hard copy plot is now drawn correctly if abs(lower Y value) is 
   greater than abs(upper Y value) and if the upper Y value is greater 
   than zero.

*  If the X-axis is logarithmic and X-axis is drawn on the Y-axis zero 
   (where the Y-axis spans zero), then the axis-crossing tick label is now 
   omitted.

*  Marquardt fitting now functions  properly if the fitting limits are reset 
   to positions which give rise to very large standard deviations.

*  Negative values in ATF files are now handled properly.

*  Negative values in ATF files are now handled properly if logarithmic 
   X-axis display or logarithmic binning is selected.  Note, however, 
   that in either case negative values are necessarily ignored.



The following bugs were fixed in PSTAT 6.0.2
============================================

*  Normalization of non-latency histograms from latency files now works
   correctly.
   
*  Sorting of single column ATF file data now works correctly.

*  Editing single column ATF file data is no longer allowed.
   
*  Square root of the bin counts with single column ATF file data is now
   drawn correctly in the hardcopy plot.

*  The standard error of P(open) is now calculated correctly.

*  Histogram axes from histogram displays called from the spreadsheet are
   now drawn correctly.
   
*  P(open) times over intervals which are less than the longest dwell
   time are now calculated properly. Previously the P(open) for these
   intervals were reported as "0" rather than "1" although the channel
   was continuously open.

*  SetLimitsToMax (Analyze/Limits) now works correctly. Data are not re-read
   unless the limits have been set previously.
   
*  Short event rejection is now available for IPROC files.

*  Data limits are now set properly in Parameters/leVelSetup if the low data
   limits is not zero and the upper data limit is set to zero.
   
*  Cursor-selected data limits (Analyze/Limits) now works correctly with
   latency files.
   
*  Truncated last events in latency episodes are now excluded from the
   dwell-time histograms.

*  Gaussian seed estimation now works correctly with ATF files. Sporadic
   problems with Gaussian fitting should no longer occur.
   
*  Latency data can now be processed in `Binned Only' mode.

*  If data limit exceeded and `Process' was selected then the single event
   immediately above the limit was not binned in the histogram. This has been
   fixed.

*  pSTAT now reads IPROC files correctly.

*  The time for the level in `Analyze Information' is now reported correctly
   when two files are concatenated. Outlier sum, standard deviation of the
   mean, total record time from file, binned data time for level,
   sum of unbinned event times and number of events in file are now
   reported correctly in `Binned Only' mode.
  
*  All-level amplitude now works properly in `Binned Only' mode.
   
*  Non-ATF files no longer display an empty histogram when they are read after
   an ATF file.

*  If P(open) interval is too large then pSTAT recovers from the entry.
   Previously pSTAT kept entering an error loop and would not allow the
   P(open) interval to be set to an acceptable value.

*  Standard deviation (Analyze Information) is now reported correctly.
   
*  An error message is now given if Gaussian fitting is attempted with
   log-binned data from an ATF file.
   file) 

*  Fixed fitting parameters now work correctly.  Previously "fixed" parameters
   were changed during the automatic parameter adjustment routine.  Note,
   however, that the values of the fixed parameters "P" (proportions) will
   still vary as these parameters are normalized before being displayed.
   Therefore, if some Ps are allowed to change then the fixed P values will
   also appear to have changed after normalization even though they were not
   allowed to change during the fitting routine.
   
*  pSTAT no longer crashes when dislaying fitted data with the Y-axis
   set to a logarithmic display.
   
*  Binned percent of total time is now reported correctly when using
   the large memory model (10,000 events).
   
===============================================================================
   
The following bugs were fixed in PSTAT 6.0.3

*  If Amplitude data type was selected (Analyze/Type), and if the level was
   set higher than that which was available in the next data file, then the
   amplitude level could not be selected for the new file. This has been fixed.
   
*  "Number of bins in fit" is now displayed with the fitting statistics.

*  Sum of squared errors was rounded of when computing the model correlation
   statistic.  This gave rise to slightly smaller values for the T-statistic
   than should have been the case. This has been corrected.
   
*  Selecting automatic interval setting for P(open) data no longer resets
   the 'P(open) to level' and 'Number of channels' in patch to the default
   values.
   
*  Burst analysis now records first closings and first openings in bursts
   correctly.

*  pSTAT no longer crashes when very large data files are read in.  This
   problem arose only when the amplitude histogram bin width was small.
   
*  pSTAT no longer crashes when the "all level amplitude including level zero"
   data type is requested.
   
*  The Gaussian mean is now correctly reported in the Analyze/Information
   screen (previously this value was reported as "0").
   
*  pSTAT now retains the drive information if a root directory (e.g. "D:\")
   is specified. Previously, pSTAT would default to the home directory.

*  For Gaussian data, pSTAT no longer automatically sets the upper limit of
   the X-axis histogram display to zero for negative amplitude data.  This
   allows for a correct representation of the actual histogram range (with
   the maximum of 256 bin in effect) if the amplitude values are widely
   scattered.

*  pSTAT now reads a single column ATF file completely.

===============================================================================
   
The following bugs were fixed in PSTAT 6.0.4

*  The derivative for log-binned data was not derived correctly.

*  P(open) works properly when "ALL" selected for "level".

*  Display of information in Analyze/Information is now correct.

*  Latency minimum duration window now works properly.

*  Added bin mode bins per decade option to latency options list.

*  Graph now displays "Log Latency" (ms) if latency data are log binned.

===============================================================================
