How do I get started? I've worked through Marc's nice book, did some of the tutorials, and have loaded SIPmath modeler tools 3.2.6 in my Excel.

Now I'd like to take a dataset (column of 20-30 numbers) that represents historical occurrences, resample to 1000, then show histograms and percentiles. Then I'll do the same for other, related datasets that are likely entangled.

Once I get the hang of it, I can work from there: any help with walking me through the initial steps would be appreciated!

Now I'd like to take a dataset (column of 20-30 numbers) that represents historical occurrences, resample to 1000, then show histograms and percentiles. Then I'll do the same for other, related datasets that are likely entangled.

Once I get the hang of it, I can work from there: any help with walking me through the initial steps would be appreciated!

Hello all,

I have just been introduced to the idea of SIP math. I am a statistician in the business world and thus would like to learn more. I had a question perhaps some one could point me in the right direction. I would like to update a SIP with real data. That is I would like to replace some of my simulated data with observed data. I am trying to find a source to prove what I want to do is mathematically sound, and how I would choose which simulated observations to switch out with the real data. Was just wondering if anyone has done something like this or would know where to look for how to validate a technique like this.

Thank you

I have just been introduced to the idea of SIP math. I am a statistician in the business world and thus would like to learn more. I had a question perhaps some one could point me in the right direction. I would like to update a SIP with real data. That is I would like to replace some of my simulated data with observed data. I am trying to find a source to prove what I want to do is mathematically sound, and how I would choose which simulated observations to switch out with the real data. Was just wondering if anyone has done something like this or would know where to look for how to validate a technique like this.

Thank you

Post has attachment

**SIPmath Meteorology**

"To get a better sense of whether or not the severe storms' lack of landfall was an outlier or not, they looked at historical records and ran computer simulations of hurricane seasons from 1950 to 2002. They ran the simulations 1,000 times and estimated that a 9-year stretch like this is likely to occur only once in about 177 years."

Post has attachment

**Simpson's Paradox**--a simple description.

Post has attachment

**SIPmath 2.0 Standard from ProbabilityManagement.org**

ProbabilityManagement.org's open SIPmath 2.0 standard links together risk models in Excel, Crystal Ball,@RISK, Risk Solver, XLSim and many other environments.

Free tools for creating interactive simulation models in Excel, models that don't need macros or add-ins to run, are available from our web site. These can be especially useful to Crystal Ball and @Risk users; they can get interactivity by running their models into an interactive SIPmath environment.

For more information, visit

http://probabilitymanagement.org/tools.html

A SIP is a unit simulator. Trial number in, data out.

Post has attachment

A good Article by David Vose explaining some of the hazards of calculating uncertainty without SIP math.

Post has attachment

**SIP math vs Monte Carlo**

This is a topic that should get more attention.

#probabilitymanagement #montecarlo #simulation

Post has attachment

**DistShaper6 does standard SIPs in XML**

I've upgraded DistShaper to put out the proposed XML standard for SIPs. See http://smpro.ca/pjs/distshaper .

Use the "CSV coded SIP" format.

**Red Word Translator**

In

*The Flaw of Averages*, Sam introduced the idea of

**Red Words**. I think it would be useful to extend that tradition in the context of SIPs and SIP math. Feel free to post corrections and additions.

**Conditional Probability Distribution**

A filtered SLURP using the sample values of one of the included SIPs to provide the filter. Conventionally, that's

*F(i) = (a < SIPx(i) < b)*

but it can be any boolean function.

**Continuous Random Variable**

Discrete Random Variable - computers don't do continuous.

**Correlation**

Scatterplot of two SIPs.

**Cumulative Distribution Function**

A SIP's percentile curve. A plot of value vs rank.

**Discrete Random Variable**

An uncertain variable or SIP.

**Joint Probability Distribution**

The SIP result of a calculation with two or more SIPs

**Joint Probability Function**

Arithmetic - The SIP math used to calculate with two or more SIPs.

**Probability Density Function**

A histogram of a SIP. A SIP's Shape.

**Probability of X**

The fraction of a SIP's elements for which X is true.

**Regression**

Taking a richly-detailed SIP and replacing it with a featureless line.

**Standard Deviation**

Not defined for most SIPs

**Utility Theory**

Risk Attitude. Do you feel lucky?

**Variance**

The square of not defined for most SIPs

**Variate**

Uncertain Variable. Usage is restricted to people with PhDs.

Wait while more posts are being loaded