Content Variance

Hi,

is there a possibility to measure the content variance as object statistic variable (in the experimenter)?

I do not mean the variance of several mean content values of the associated replications.

I mean the variance of the content of within one single replication like the mean of the content.

Thanks!

Congshi

You don´t talk about the average content of a queue after one replication, don´t you? Because this you could find in the "Perfomance Measures".
What you are looking for, is a chart about the variance of the content, aren´t you?

Exactly! Anthony has posted a similar formula for the staytime variance in this thread: http://www.flexsim.com/community/for...read.php?t=644, but I have no idea how to transform this formula to measure the content variance.

Thanks!

So you want to show some graphs about the different replications, but all in one chart? And the graphs should show the same as the content chart in the statistic tab?

No, I want to have a Performance Measure in the Experimenter just like average content of the single replications and its mean and sd values, just for the variance of the content.

Ich möchte die Ergebnisse der einzelnen Szenarien sowie deren Versuchswiederholungen nicht nur für die Mittelwert der jeweiligen Versuchswiederholungen haben, sondern auf die Varianzen. Dabei soll der Experimenter die Varianz des Bestands für jede einzelne Versuchswiederholung messen und daraus den Mittelwert und Standardabweichung für alle Wiederholungen eines Szenarios bilden. Ich hätte dann im PFM Fenster eine Liste von z.B. 20 Szenarien, zu denen der jeweilige Mittelwert, Konfidenzintervall und Standardabweichung sowie MinMax-WErte aus z.B. 20 Wiederholungen je Szenario gebildet und anschließend auch angezeigt wird.

Also keine Graphen, sondern nur Werte.

Thanks

Now I understood, but purely can´t help you. I´m sure there will be a problem with getting the variance of average content because at beginning of a scenario the content is always 0 and this will interrupt your values.

If setting the content at the beginning of the szenario with the value of the average content, you maybe can solve this. But probably this isn´t a good solution.

Just thinking about the warm up time in the experimenter!?

Another idea is to get the minimum and maximum content and subtract it from average content. The higher number could be nearly what you are looking for.

But to be honest, I hope anybody else got the idea you need.

I even understand your problem with the formula for the variance of staying time. Because stayingtime is writen only once for every item. The content will grow up slowly and then rise or fall about the time.