Laserfiche WebLink
Commands TSTool Documentation <br />Commands - 6 <br />supported for some input types. TSTool displays and o utput products indicate missing data as blanks , by <br />showing the missing data value, or a string (e.g., NC ), depending on the constraints of the product. For <br />example, an HTML time series highlights missing values and shows flags as a superscript . <br /> <br />Filled ti me series often are required for use in computer models. TSTool provides a number of features <br />to fill time series data. The data filling process consists of analyzing available data and using the results <br />to estimate missing data values. The estimation process can be simple or complex, resulting in varying <br />degrees of estimation error and statistical characteristics of the final time series. The data analysis uses <br />data that are available at the time that the fill command is encountered. Consequently if values have been <br />changed since the initial read (e.g., because of layered fill commands), the changed values may impact the <br />analysis. Basic statistical properties of the original data are saved after the initial read to allow use in later <br />fill commands. For example, for monthly time series, the historical monthly averages are computed after <br />the initial read to allow use with a F illHistMonthAverage() command. Fill commands often <br />provide a FillFlag parameter, which allows filled values to be annotated. Th e flags can then be <br />displayed in reports and graphs. <br /> <br />The overall period that is being filled is controlled by the time series period or analysis period that is <br />specified with fill commands. TSTool will not automatically extend the period of a filled tim e series <br />after the time series is initially read. Use the S et Input Period() and S etOutputPeriod() <br />commands to control the time series period. <br /> <br />The following table lists the fill techniques that are supported by TSTool. <br /> <br />TSTool Fill Techniques and Associate d Commands <br /> <br />Technique Command Typical Use <br />Constant F illConstant() Use when missing data can be estimated as a <br />constant. For example, if only the early period of a <br />"regulated" (e.g., reservoir) time series is missing, it <br />may be appropriate to set the valu es to zero. <br />Monthly total, <br />daily pattern <br />F illDayTSFrom <br />2MonthTSAnd1DayTS() Use to estimate a daily time series by applying the <br />pattern of a related daily time series to monthly <br />totals from the related and current time series. For <br />example, use to estimate daily streamflow from <br />monthly total values. <br />Fill from time <br />series <br />F illFromTS() Use non -missing values from a time series to fill <br />missing values in another time series. <br />Historic al <br />Monthly <br />Average <br />F illHist <br />MonthAverage() Use with monthly time series to es timate missing <br />monthly values as the average of historic monthly <br />values. For example, if applied to monthly <br />precipitation data, a missing July value would be set <br />to the average of observed July precipitation values <br />(zero is an observation). <br />Historical Ye ar <br />Average <br />FillHist <br />YearAverage() Use with yearly time series to estimate missing data <br />as the average of annual values. <br />Interpolation FillInterpolate() Use to estimate missing data by interpolating <br />between non -missing values. For example, use to <br />estimate reservoir level changes. <br />Mixed Station FillMixedStation() This command tries various combinations of <br />FillRegression() and FillMove2() <br />60