How Can I Generate 10,000 Sample Values from a @RISK Distribution?

Question:
I want to generate 10,000 sample values from a distribution, e.g., RiskLogLogistic(-0.0044898, 0.045333, 2.5862). How can I do this?

Answer:
You have five methods, split into simulation and non-simulation approaches.

Simulation Methods

1. Run a simulation:

   • Create a workbook with the distribution in a cell.

   • Set iterations to the desired number of samples and run the simulation.

2. Extract simulation data:

   • Option A: Click the x-subscript-i icon → Simulation Data window → Right-click column → Copy → Paste into Excel.

   • Option B: Excel Reports → Simulation Data (Inputs) or Simulation Data → new sheet created with sampled values.

   • Option C: Browse Results → select cell → change Statistics grid to Data Grid → Right-click column heading → Copy → Paste into Excel.

Notes:

• These methods use your Simulation Settings sampling type (default: Latin Hypercube, which better matches percentiles).

• Correlations are honored during simulation.

Non-Simulation Methods

1. VBA Sampling (Industrial/Professional):

   • Use the Risk.Sample method to loop and generate samples.

   • See: Sampling @RISK Distributions in VBA Code.

2. Random/Static Dice Icon:

   • Activate the icon, enter the distribution in a cell, then click-and-drag to create multiple samples.

   • Copy and Paste Special → Values to save the sampled numbers.

Notes:

• Non-simulation methods always use Monte Carlo sampling.

• Correlations are not applied in non-simulation sampling.

Summary:

• Simulation methods: Use Latin Hypercube, honor correlations, better percentile matching.

• Non-simulation methods: Faster, ignore correlations, use Monte Carlo sampling.

 

Related Articles:

better than traditional Monte Carlo sampling 

 Sampling @RISK Distributions in VBA Code