DUMPStat patches

The following table contains current DUMPStat patches. You can click on the description hyperlink for more details on how each patch will affect your program. Click once on the file name to download the file, and save it to your DUMPStat program directory.

Note that DUMPStat 2.1.1 must be installed prior to installing the cumulative maintenance patch (DUMP218.EXE). Do not try to install the patch over a 2.0 or 2.1 installation — this will cause DUMPStat to stop functioning.

ISG.EXE is a password-protected file. After saving it to your DUMPStat program directory, click on Statistical Guide from DUMPStat's Help Menu.

DUMP218.EXE and DS_USER.EXE are self-extracting files. Follow these steps to install them:

1. Double-click on the downloaded file to extract its contents.
2. Choose to overwrite the existing file(s) when prompted. Click on OK. (Note: if the overwrite warning does not appear, you may be extracting the patch files to the wrong directory.)
3. Do not delete the self-extracting file once you have unzipped it — this file can be a useful reference for tech support if you experience any problems later on.
4. If you have multiple installations of DUMPStat you will need to ensure that the self-extracting file is unzipped to each DUMPStat directory.

DUMPStat 2.1.8 Statistical Guide Installation
This is an installation file for the DUMPStat Statistical Guide PDF. If you cannot access the Statistical Guide from DUMPStat's Help Menu, save this file to your DUMPStat program directory. In DUMPStat, click on Statistical Guide from the Help Menu.
Note that the Statistical Guide PDF can be accessed only from a fully registered DUMPStat.  

DUMPStat 2.1.8 maintenance release
This is a self-extracting file containing DUMPStat program and statistical components, as well as accessory files and an updated Help file.

April 25, 2003 changes:

• The January 2003 change to outlier detection in Up vs. Down Prediction Limits introduced an error such that samples for the last well in the list of upgradient wells were not compared to the median value. Since only Dixon's test was being used to determine outliers, the last well could have samples declared as outliers without them being greater than 10 times the median value which was inconsistent. This has been corrected.   

April 17, 2003 changes:

• Manual reporting limits were not implemented correctly in Intra-Well Sublist Prediction Limits such that nondetects were ignored in the calculation of background limits. This did not affect the nonsublist version nor were any other analyses affected. The error was introduced in the January update.
• Minor enhancements were made to the new site lock disabling feature.
• Inaccessible archival directories no longer cause a site to be permanently set to "read only" status. Previously, if a network drive were unavailable when DUMPStat was started a site could be put in a "read only" state such that the site could not be saved. Now the site is set to "read only" only during the time when the drive is not accessible and access is reset to "read write" when the program starts with the archival directory available to the computer.  

January 2003 changes:

• The outlier test was altered to eliminate round-off error on internal processing. For a value to be considered an outlier it must be classified as such by Dixon's test and be either 3 or 10 times the median value depending on the analysis being performed. Prior to this correction values that were exactly 3 (or 10) times the median value may have been allowed to be considered outliers. Now values must be strictly greater than 3 (or 10) times the median value in addition to Dixon's test to be classified as outliers.
• When manual reporting limits are enabled, they are still used in the calculations but are no longer plotted on the graphs. Instead, the nondetect samples are plotted at their original merged values, and the manual reporting limit for that constituent is instead indicated on the graph by a small diamond. If you are doing confidence limits, surface water or air analyses, the manual reporting limit is divided by two. In cases where manual reporting limits are not enabled, or where they are but the constituent doesn't have one, the median [lab] reporting limit, or half the limit, is used.
• Surface water and air nondetects used to be plotted as half the median reporting limit instead of half the original lab values. This was inconsistent with nondetects in confidence limits, which were plotted at half their lab values. This has been corrected now so that confidence limits, surface water, and air analyses all have nondetects plotted at their original merged value/2.
• There is a new display option for the statistical power charts to overlay five dotted lines on the graph. Two vertical lines appear at 3 and 4 SDs (standard deviations) along the horizontal axis, and run from the top to the bottom of the area bounded by the axes. Three horizontal lines appear at 5%, 50%, and 80% likelihood on the vertical axis. The lines provide a visual reference for judging the position and steepness of the power curve, and are displayed either all together or not at all. They will also be printed if you have them turned on.
• DUMPStat now allows more flexibility in site locking. You may turn off site locking altogether, and you may also choose whether to have DUMPStat save to the archival site directory.
• Previously, DUMPStat observed only manual outliers while testing for trends in Time Series (if manual outliers were enabled). Now, DUMPStat screens for automatic outliers too, and then performs the trend test, if you are identifying historical trends.
• When you select a data file to merge, text files are now listed as .PRN as well as .TXT. in the ‘List Files of Type’ box. The recognized file types list includes .rec (leftover records), .csv, and .dat, and all files.
• When DUMPStat and Explorer are using the same site, there used to be conflicts when each program was attempting to access the database. There is now a higher degree of compatibility between the two programs, and there will be no failures if the database is being accessed simultaneously.
• For setting manual reporting limits and standards (in Select for Assessment), if there was at least one item that did not have units (like pH), you were asked whether to save your changes if you simply canceled out of the dialog without changing anything. Now, there is no enquiring message box if you cancel out of the window with no changes.  

June 2002 changes:

• Previously, if you changed the printer in either DUMPStat or any of the viewers, it was necessary to shut down the program and restart it for the change to take effect. Now, the Graph, Worksheet, and Power Chart viewers allow you to make a printer switch and see the change immediately, without restarting. However, it is usually still necessary to restart DUMPStat or the Table Viewer if you have changed the printer in either one of these programs.
• When the data for a graph spans a very short period, the default time display unit in the Graph Viewer is half a year. We used to display half years as January to June. It has now been corrected to display January - July.
• Small text on analysis output has been clarified and is now easier to read.  

October 2001 changes:

• All settings chosen in the Set Time Windows dialog are now saved with zones. This includes the option to include the most recent data for Up vs. Down analyses, and to ignore all data beyond a specified month.
• All statistical options are now saved with zones.
• Zone files can now be successfully transferred between different versions of DUMPStat.
• Analytical Data Summary tables can now be custom numbered, either consecutively or all the same number.
• Custom numbering for all tables can be up to 3 digits.
• Table 6 in Up vs. Down Prediction Limits now accepts changes in title or table number.
• New default file extension (*.txt) for selecting data files to merge.
• DUMPStat communicates more smoothly with the viewer applications now, eliminating occasional error messages generated when statistical settings were changed with old output still open.
• Merge Data window more robust: no more failure due to illegally grouped numbers resulting from accidental inclusions in the wrong columns.  

April 2001 changes:

• Instances of page numbering starting at page 2 have been eliminated.
• Power chart curves now display correctly when the y-intercept is 90 or higher.   

March 2001 changes:

• The Fix Horizontal Axis function has been fixed to work with graphs that don't have any data points.
• Statistical Power Charts that couldn't be computed are flagged with a more informative message than previously.
• Log scaling has been improved so that lower prediction limits are now drawn at 0.
• Print command is handled more gracefully now if there are no graphs to view.
• New statistical option allows you to set the Alpha value as either the calculated value or the minimum of the calculated value and 0.01. This affects Up vs. Down, Intra-Well Control Charts, and Intra-Well Prediction Limits.
• No more "installation out of sync" errors resulting from unexpected Windows shutdowns.
• Outliers in Up vs. Down Prediction Limits are now accurately identified.
• For Intra-Well (& Sublist) Control Charts, Time Series, Intra-Well (& Sublist) Prediction Limits, and Confidence Limits, analyses will not fail if the "Ignore Data After" command has been set to exclude all data from the analysis.  

Statistical changes:

• In Sublist Control Charts and Prediction Limits, two-sided limits for all constituents after pH have been eliminated and the correct single limits are now being displayed.
• Outliers in Up vs. Down Prediction Limits are now accurately identified.
• For Intra-Well (& Sublist) Control Charts, Time Series, Intra-Well (& Sublist) Prediction Limits, and Confidence Limits, analyses will not fail if the "Ignore Data After" command has been set to exclude all data from the analysis.  

DUMPStat 2.1.8 User's Guide
This is a self-extracting file containing the updated DUMPStat User's Guide PDF. The User's Guide PDF can be accessed on DUMPStat's Help Menu.