Applies to: @RISK 5.x and Up

 

Getting statistics for part of a distribution without changing the simulation

If you want to simulate the full input distribution but compute statistics on only part of it, don’t truncate the input itself. Instead, apply truncation inside the statistic function.

Example:

• A1: =RiskNormal(100,10) (full distribution is simulated)

• A2: =RiskMean(A1, RiskTruncate(95)) → mean of values ≥ 95

• A3: =RiskMean(A1, RiskTruncateP(0.8,1)) → mean of top 20%

You can use RiskTruncate() (by values) or RiskTruncateP() (by percentiles) inside RiskMean, RiskStdDev, RiskPercentile, etc.
This also works when filters are applied, letting you compute stats on just the filtered subset.

Note:

• With RiskMean (simulated), stats are based on simulation data.

• With RiskTheoMean, @RISK runs a small internal simulation to approximate truncated theoretical stats.

 

Related Article:

Cutting Off a Distribution at Left or Right

Attachments:
KB449_RiskTruncateExample.xlsx