Laserfiche WebLink
Finally, as part of the statistics program (if selected by the user) any missing data that was not <br />filled by the MSM is then filled with average data. The GUI then replaces the filled files (*.xbf) <br />with the average filled data and will list years that data was filled with averages in the header <br />information of the file. In order to allow the user to create files of filled baseflow data for from <br />just the MSM and also with average monthly data filling the remaining values the default is to <br />create two types of xbf files. The original MSM output has a default name of *-l.xbf files and <br />for the files with all data filled by the MSM then the average monthly data the default name is <br />*.xbf. <br />Program Installation <br />The Mixed Station Graphical User Interface can be installed on Windows NT or Windows 95/98 <br />computers. The package is initially in aself-extracting "zip" format, which can be expanded by <br />executing the file "MixStation.exe". This creates three new files in a user specified directory (the <br />default directory is C:\Unzipped\). Executing the "setup.exe" file will initiate the installation <br />routine. <br />Use Of GUI <br />To open a file, click the "Open" button or choose "Open" from the File menu. The unfilled <br />baseflow data has been included with the graphical interface under <br />"C:\CDSS\MixStation\Data\Unfilled\" or whatever path the user specified during the installation. <br />The Mixed Station Model provides several options to the user as shown in Fi ure 1. These <br />options will not become active until at least one file has been opened. Three choices exist for the <br />type of regression analysis to be performed on the data set: Simple Linear Regression, Move.l, <br />and Move.2. The default is Simple Linear Regression and should be used in most cases. <br />Cyclicity indicates whether the cyclic or non-cyclic value is to be used to fill empty data. SEP <br />Dependent (Standard Error of Prediction Dependent) indicates that the best value, either cyclic or <br />non-cyclic, will be used to fill data. SEP Dependent is the default value and is recommended. <br />Dependent Stations indicate the number of stations in the data set. The Mixed Station Model can <br />be set up to look through all available stations to fill data (Search all Stations) or to use a limited <br />number of Independent Stations (Use Designated Stations). The default value is Search All <br />Stations and is recommended. The beginning year may also be set so that a smaller period of <br />record is filled. If the period of record is reduced, no data from before that date will be used to <br />fill in missing data. The ending year is always set to the last year of record in data set. <br />Minimum Concurrent Values indicate the minimum number of concurrent values of dependent <br />and independent variables that will be needed to consider the potential independent as a <br />candidate for use in the filling process. Mixed Station also allows the user to specify a <br />confidence interval for filling missing data. The default confidence interval is 95% and is <br />recommended. <br />Once all options have been set (the default options should work for most users) clicking the <br />"Run" button or choosing "Run" from the File menu will execute the Mixed Station Model <br />program, which will appear as a DOS box on your screen. <br />Appendix E E-114 <br />