https://sourceforge.net/p/freemuse/wiki/Calculation%2520section/Recent changes to Calculation sectionenMon, 06 Jun 2011 18:58:42 -0000https://sourceforge.net/p/freemuse/wiki/Calculation%2520section/<pre></pre>Marco WolfMon, 06 Jun 2011 18:58:42 -0000https://sourceforge.netdbe12e06500704276449875d44233a6c2040d67chttps://sourceforge.net/p/freemuse/wiki/Calculation%2520section/<pre>--- v1 +++ v2 @@ -25,85 +25,79 @@ Attributes for calculation section --- The following attributes can be set in the calculation section: -||<bgcolor="#999999" style="vertical-align: top;">Attribut ||<bgcolor="#999999" style="vertical-align: top;">Posible values ||<bgcolor="#999999" style="vertical-align: top;">Description || -||mcsimulations ||(#|a) ||Set number of MC simulatins (a - adaptive, # - number * 10000) || -||rngenerator || (__'mt__'|wh) || Uniform Random number generator to be used (mt- Mersenne Twister, wh - Wichmann-Hill) || -||dimensions || # ||number of output dimenstions. Negative numbers are complex valued quantities || -||staticseed ||(on|__'off__') || If set to __on__ the RNG is initialized with the same value all the time || -||formulaformat ||(__'infix__'|postfis) ||Formula format of the whole calculation || -||outputdigits || # ||Number of digits to be stored in the output file || - +[[img src=csheader.jpg alt=foobar width=100%]] + See the section [adaptive MC](Adaptive Monte Carlo) for setting the attributes for adaptive MC. Variables and variation loops === [Variables] and variation loops are used to set global parameters used in the calculation of the measurement uncertainty. They can be used for shared values in processes and instances. Variables --- As variables can be defined in the calculation section as well as in the [process section](Process section), we refer to the section [Variables]. Simple for-loop --- Often you want to compare different outputs with different values for a parameter/variable or you want to define sequential measurements. In __MUSE__ you use variation variables for that. ~~~~ <variation name="varname" from="1" to="10" step="2"/> ~~~~ With this simple line you will simulate five different scenarios. Each variation represents one value of the variable __varname__. The values for __varname__ will be (1, 3, 5, 7, 9). Variation list --- Sometimes you are not interested in fixed interval variations, but you want to define for example when your different measurements take place. Therefore we have a construct called variation list, which lets you define exactly this scenario. ~~~~ <variationlist name="varname"> <value>8.0</value> <value>8.2</value> <value>8.5</value> <value>9.0</value> </variationlist> ~~~~ This can be an abstract definition of time when you did the measurements for 8:00, 8:12, 8:30 and 9:00. Variation set --- The variation list allows you to define almost any scenario. But it can be annoying if you have to use one variation variable, when you want to use a set of different variation variables and think about formulas how to express the dependence of such variables. What you can use in this case are the variation sets. They allow you to define different sets of variables for example for different dates. ~~~~ <variationset name="varset"> <set> <variable name="time"><value>8.0</value></variable> <variable name="temp"><value>18.0</value></variable> <variable name="staff"><value>0.4</value></variable> </set> ... <set> <variable name="time"><value>9.0</value></variable> <variable name="temp"><value>20.0</value></variable> <variable name="staff"><value>0.9</value></variable> </set> </variationset> ~~~~ You can interpret the example like this: at 8:00am you have a temperature of 18°C in the laboratory, the staff is still a bit sleepy and therefore gets just a reliability factor of 0.4; at 9:00am you do your last measurement, the temperature rises to 20°C and the staff is nearly at full concentration with a reliability factor of 0.9. The big advantage of this kind of definition of variation sets is, that you have all your information compressed at one point and the comparison between different sets should be very easy. Related Topics === Please refer to this pages for more information: * [Scope] == More advanced options in calculation section == There are more advanced options that can be set and used in the calculation section. Please refer to the following pages for more information. Settings: * [Adaptive Monte Carlo] * [Random number generator] Parameter sections: * [Fade out of parts of the calculation](Fade out) - * [Analyzing_results] + * [Analyzing results] * [Validating before simulating](Validation) * [Defining and working with measurement series](Measurement series) </pre>Marco WolfMon, 06 Jun 2011 18:58:19 -0000https://sourceforge.net22ba968d029bbbad1be0f92f0bb234a49e8d613ahttps://sourceforge.net/p/freemuse/wiki/Calculation%2520section/Calculation section === In the calculation section you define everything that belongs to the calculation like the number of Monte Carlo simulations, the type of random number generator to be used, the number of output digits and so on. Beside the parameters for the calculation you define here the summarizing formulas for your measurend(s) of your simulation. __MUSE__ is able to handle multivariate output by defining more than one output formula. The formulas are encapsulated in an element called __measurand__. A very simple example with one output formula is the following. ~~~~ <calculation mcsimulations="100"> ... <measurand>g1 + g2</measurand> </calculation> ~~~~ Here the variables __g1__ and __g2__ are added. The result is a one dimensional output file. A more Advanced example follows where we have three output quantities. ~~~~ <calculation mcsimulations="100"> ... <measurand> <formula name="s1">m - mt / V1 + V2</formula> <formula name="s2">m + mt / V3 + V4</formula> <formula name="s3">m + mt / V5 + V6</formula> </measurand> </calculation> ~~~~ The calculation section above defines three formulas named __s1__, __s2__ and __s3__. The formulas are defined in infix notation. __MUSE__ transforms this formulas into postfix notation internally. To avoid this conversion you can define the formulas directely in postfix notation. For details see the section [formula format](Formula format) of __MUSE__. __MUSE__ comes with several mathematical functions like the square root, trigonometrical functions and also some predefined constants like pi. For details see the list of [keywords](Keywords). Attributes for calculation section --- The following attributes can be set in the calculation section: ||<bgcolor="#999999" style="vertical-align: top;">Attribut ||<bgcolor="#999999" style="vertical-align: top;">Posible values ||<bgcolor="#999999" style="vertical-align: top;">Description || ||mcsimulations ||(#|a) ||Set number of MC simulatins (a - adaptive, # - number * 10000) || ||rngenerator || (__'mt__'|wh) || Uniform Random number generator to be used (mt- Mersenne Twister, wh - Wichmann-Hill) || ||dimensions || # ||number of output dimenstions. Negative numbers are complex valued quantities || ||staticseed ||(on|__'off__') || If set to __on__ the RNG is initialized with the same value all the time || ||formulaformat ||(__'infix__'|postfis) ||Formula format of the whole calculation || ||outputdigits || # ||Number of digits to be stored in the output file || See the section [adaptive MC](Adaptive Monte Carlo) for setting the attributes for adaptive MC. Variables and variation loops === [Variables] and variation loops are used to set global parameters used in the calculation of the measurement uncertainty. They can be used for shared values in processes and instances. Variables --- As variables can be defined in the calculation section as well as in the [process section](Process section), we refer to the section [Variables]. Simple for-loop --- Often you want to compare different outputs with different values for a parameter/variable or you want to define sequential measurements. In __MUSE__ you use variation variables for that. ~~~~ <variation name="varname" from="1" to="10" step="2"/> ~~~~ With this simple line you will simulate five different scenarios. Each variation represents one value of the variable __varname__. The values for __varname__ will be (1, 3, 5, 7, 9). Variation list --- Sometimes you are not interested in fixed interval variations, but you want to define for example when your different measurements take place. Therefore we have a construct called variation list, which lets you define exactly this scenario. ~~~~ <variationlist name="varname"> <value>8.0</value> <value>8.2</value> <value>8.5</value> <value>9.0</value> </variationlist> ~~~~ This can be an abstract definition of time when you did the measurements for 8:00, 8:12, 8:30 and 9:00. Variation set --- The variation list allows you to define almost any scenario. But it can be annoying if you have to use one variation variable, when you want to use a set of different variation variables and think about formulas how to express the dependence of such variables. What you can use in this case are the variation sets. They allow you to define different sets of variables for example for different dates. ~~~~ <variationset name="varset"> <set> <variable name="time"><value>8.0</value></variable> <variable name="temp"><value>18.0</value></variable> <variable name="staff"><value>0.4</value></variable> </set> ... <set> <variable name="time"><value>9.0</value></variable> <variable name="temp"><value>20.0</value></variable> <variable name="staff"><value>0.9</value></variable> </set> </variationset> ~~~~ You can interpret the example like this: at 8:00am you have a temperature of 18°C in the laboratory, the staff is still a bit sleepy and therefore gets just a reliability factor of 0.4; at 9:00am you do your last measurement, the temperature rises to 20°C and the staff is nearly at full concentration with a reliability factor of 0.9. The big advantage of this kind of definition of variation sets is, that you have all your information compressed at one point and the comparison between different sets should be very easy. Related Topics === Please refer to this pages for more information: * [Scope] == More advanced options in calculation section == There are more advanced options that can be set and used in the calculation section. Please refer to the following pages for more information. Settings: * [Adaptive Monte Carlo] * [Random number generator] Parameter sections: * [Fade out of parts of the calculation](Fade out) * [Analyzing_results] * [Validating before simulating](Validation) * [Defining and working with measurement series](Measurement series) Marco WolfMon, 06 Jun 2011 18:56:23 -0000https://sourceforge.net0da7a76560a9f14f849e3fce7a12104218ad79f0