Difference between revisions of "Tutorial: Reviewing a model"

Line 27: Line 27:
 
</div>
 
</div>
  
==Recognizing influence diagrams==
+
==Recognize influence diagrams==
 
In this chapter, you will delve into some of the details of the ''Rent vs. Buy'' model. You will not use the top diagram that you used in [[Tutorial: Open a model to browse|Chapter 1]].
 
In this chapter, you will delve into some of the details of the ''Rent vs. Buy'' model. You will not use the top diagram that you used in [[Tutorial: Open a model to browse|Chapter 1]].
  
Line 44: Line 44:
 
:[[File:Chapter 2.3.png]]
 
:[[File:Chapter 2.3.png]]
  
==Opening Object windows==
+
==Open Object windows==
 
Every object in Analytica has an associated [[Object window]] containing detailed information about it. You can display the '''Object''' window of any variable by double-clicking its node in the [[Diagram window|influence diagram]].
 
Every object in Analytica has an associated [[Object window]] containing detailed information about it. You can display the '''Object''' window of any variable by double-clicking its node in the [[Diagram window|influence diagram]].
  
Line 55: Line 55:
 
<tip>You can enter numbers with a suffix abbreviation, so Buying Price can be defined as either 140K or 140000. A quick reference for these suffixes is given on [[Number formats]] page.</tip>
 
<tip>You can enter numbers with a suffix abbreviation, so Buying Price can be defined as either 140K or 140000. A quick reference for these suffixes is given on [[Number formats]] page.</tip>
  
==Moving between Object windows==
+
==Move between Object windows==
 
You have opened the [[Object window]] of a variable (''Buying price'') by double-clicking its node in the [[Diagram window|influence diagram]].
 
You have opened the [[Object window]] of a variable (''Buying price'') by double-clicking its node in the [[Diagram window|influence diagram]].
  
Line 80: Line 80:
 
:[[File:Chapter 2.9.png]]
 
:[[File:Chapter 2.9.png]]
  
==Using the Attribute panel==
+
==Use the Attribute panel==
 
As an alternative to viewing a variable’s attributes in a separate window, you can inspect them in the [[Attribute panel]], which is an auxiliary window pane that you can open below the influence diagram.
 
As an alternative to viewing a variable’s attributes in a separate window, you can inspect them in the [[Attribute panel]], which is an auxiliary window pane that you can open below the influence diagram.
  
Line 93: Line 93:
 
:[[File:Chapter 2.11.png]]
 
:[[File:Chapter 2.11.png]]
  
==Inspecting definitions in the Attribute panel==
+
==Inspect definitions in the Attribute panel==
 
The [[Attribute panel]] allows you to inspect any attribute of a variable.
 
The [[Attribute panel]] allows you to inspect any attribute of a variable.
  
Line 110: Line 110:
 
In [[Tutorial: Open a model to browse|Chapter 1]] you saw that ''Rate of inflation'' is defined as a [[normal]] distribution with a [[mean]] of 3.5 and a [[Sdeviation|standard deviation]] of 1.3.
 
In [[Tutorial: Open a model to browse|Chapter 1]] you saw that ''Rate of inflation'' is defined as a [[normal]] distribution with a [[mean]] of 3.5 and a [[Sdeviation|standard deviation]] of 1.3.
  
==Opening modules==
+
==Open modules==
 
Analytica models generally contain '''''modules'''''. Each [[Modules and Libraries|module]] contains the details of a part of the model, also represented as an [[influence diagram]]. In the ''Rent vs. Buy'' model, ''Cost to Buy'' and ''Cost to Rent'' are both modules.
 
Analytica models generally contain '''''modules'''''. Each [[Modules and Libraries|module]] contains the details of a part of the model, also represented as an [[influence diagram]]. In the ''Rent vs. Buy'' model, ''Cost to Buy'' and ''Cost to Rent'' are both modules.
  
Line 148: Line 148:
 
:[[File:Chapter 2.23.png]]
 
:[[File:Chapter 2.23.png]]
  
==Inspecting values in the Attribute panel==
+
==Inspect values in the Attribute panel==
 
The [[Attribute panel]] allows you to view certain attributes, such as a variable’s '''''value''''', that are not (initially) displayed in an [[Object window]].
 
The [[Attribute panel]] allows you to view certain attributes, such as a variable’s '''''value''''', that are not (initially) displayed in an [[Object window]].
  
Line 161: Line 161:
 
It is faster to compute a mid (deterministic) value than an uncertain (probabilistic) value, so it is useful for conducting initial checks of a model before performing any uncertainty analysis.
 
It is faster to compute a mid (deterministic) value than an uncertain (probabilistic) value, so it is useful for conducting initial checks of a model before performing any uncertainty analysis.
  
==Displaying results==
+
==Display results==
 
When you are viewing a model’s [[Diagram window|influence diagram]], you can evaluate any variable and display its value in a [[Result window]].
 
When you are viewing a model’s [[Diagram window|influence diagram]], you can evaluate any variable and display its value in a [[Result window]].
  

Revision as of 17:40, 1 June 2018



Release:

4.6  •  5.0  •  5.1  •  5.2  •  5.3  •  5.4  •  5.5

This Tutorial page shows you how to explore a model by examining the Rent vs. Buy model's:

  • Influence diagrams
  • Variables
  • Attributes
  • Definitions
  • Results

This assumes you have started Analytica and have opened the Rent vs. Buy model. If this is not the case, see Opening the Rent vs. Buy model. If you are using the model as modified from Chapter 1, change the value of Time horizon back to 10, the value of Monthly rent back to 1200, and the value of Buying price back to 140K. Also change the Rate of inflation back to a normal distribution with a mean of 3.5 and a standard deviation of 1.3.


Tutorial Video: Review/explore a model (8 minutes)


Recognize influence diagrams

In this chapter, you will delve into some of the details of the Rent vs. Buy model. You will not use the top diagram that you used in Chapter 1.

Chapter 2.1.png

The details of an Analytica model display in an influence diagram window. An influence diagram (shown below) is a graphical representation of a model, showing how different variables in the model interact with each other. A typical influence diagram consists of a number of nodes connected by arrows.

Chapter 2 2b.png

Nodes represent variables and appear as boxes, ovals, hexagons, and other shapes. Different node shapes represent different types of variables. Analytica uses the term variable broadly to include anything that has a value or can be evaluated. Note that many of the variables have the same names as the inputs and output at the top diagram that you used in Chapter 1. The top diagram provides an easy way to see and change these nodes’ values.

Arrows connecting different variables indicate a relation between the variables. The arrow connecting Rate of inflation to Appreciation rate indicates that the value of the Appreciation rate variable depends on the value of the Rate of inflation variable. In the Rent vs. Buy model influence diagram, Cost to Buy depends on the Buying price, Rate of inflation, Appreciation rate, Discount rate, and Time horizon variables.

The following figure illustrates different types of nodes.

Chapter 2.3.png

Open Object windows

Every object in Analytica has an associated Object window containing detailed information about it. You can display the Object window of any variable by double-clicking its node in the influence diagram.

Chapter 2.4.png

Information about a variable is provided in a list of attributes. Attributes include the variable’s class (for example, decision, chance, or constant), identifier, units, title, description, definition, inputs, and outputs. See the illustration below.

Chapter 2.5.png
Tip
You can enter numbers with a suffix abbreviation, so Buying Price can be defined as either 140K or 140000. A quick reference for these suffixes is given on Number formats page.

Move between Object windows

You have opened the Object window of a variable (Buying price) by double-clicking its node in the influence diagram.

The Object window contains a list of the variable’s inputs and outputs, if there are any.

You can open the Object window for any input or output variable by double-clicking the one you wish to view.

Chapter 2.6.png

Analytica opens the Object window for Mortgage loan amount.

Chapter 2.7.png

Note in the figure above that the Title of Mortgage loan amount is different from the variable’s identifier, Mortgage. The title is what the model user normally sees; the identifier is used as a mathematical symbol in the definitions of other variables that depend on this variable.

The definition of the Mortgage loan amount is an expression, the sum of Buying price and Down payment (which is a negative amount). The definition refers to these variables by their identifiers.

Inputs lists the identifiers and titles of the variables in the definition. Buying price, the variable you just examined, is one of the inputs. The other input of Mortgage loan amount is Down payment.

Chapter 2.8.png

The Object window now displays the attributes of Down payment.

Chapter 2.9.png

Use the Attribute panel

As an alternative to viewing a variable’s attributes in a separate window, you can inspect them in the Attribute panel, which is an auxiliary window pane that you can open below the influence diagram.

The Attribute panel allows you to rapidly examine one attribute at a time of any variable in the model. You select the variable you wish to view and select the attribute to examine from a popup menu.

The variable Buying price should be highlighted with a title in white, indicating that it is selected; if it is not, select it by clicking it once.

Chapter 2.10.png

By default, Analytica displays the description of the selected node (e.g., Buying price) in the Attribute panel.

Chapter 2.11.png

Inspect definitions in the Attribute panel

The Attribute panel allows you to inspect any attribute of a variable.

In this section, you will see the definition of two variables that you viewed in the top-level diagram in Chapter 1.

Chapter 2.12.png
Chapter 2.13.png
Chapter 2.14.png

When a variable is defined as an uncertainty distribution, a button appears in the Definition field.

Chapter 2.15.png

In Chapter 1 you saw that Rate of inflation is defined as a normal distribution with a mean of 3.5 and a standard deviation of 1.3.

Open modules

Analytica models generally contain modules. Each module contains the details of a part of the model, also represented as an influence diagram. In the Rent vs. Buy model, Cost to Buy and Cost to Rent are both modules.

Modules can also contain other modules. In this manner, a large model with hundreds of variables can be organized into a hierarchy of modules, each small enough to be easily understood.

Chapter 2.16.png

Analytica displays the influence diagram of the Cost to Buy module. This module contains three additional modules: Out-of-pocket costs to own, Future sales proceeds, and Opportunity cost.

Chapter 2.17.png

The input arrowhead (without a trailing line) shows that the node to the right of the arrow has one or more inputs from outside this module.

The output arrowhead shows that the node to the left of the arrow has one or more outputs outside this module.

Chapter 2.18.png
Chapter 2.19.png

Analytica limits the number of open windows at each level of the model hierarchy to minimize clutter on your screen. See Managing windows in the Analytica User Guide for information on how to open more than one module Diagram window at a time.

Chapter 2.20.png

The combined arrowhead, shown above, indicates that the node has one or more inputs from outside this module, plus the input variable in this module.

You can also navigate the model by tracking a variable’s inputs or outputs.

Chapter 2.22.png

The Out-of-pocket costs to own module diagram is brought to the front, with the Insurance node selected.

Help balloons

Help balloons pop up when you hover over a node for about a second, and display the title, units, and description for a variable. These allow you to browse information very rapidly without having to open the object window. Balloons do not appear when objects do not have the Help or Description attribute filled in.

Chapter 2.24.png
Chapter 2.23.png

Inspect values in the Attribute panel

The Attribute panel allows you to view certain attributes, such as a variable’s value, that are not (initially) displayed in an Object window.

Chapter 2.25.png
Chapter 2.26.png

If Value was not previously computed, Analytica computes the variable’s value deterministically, assuming that all of the input probability distributions are fixed at their median values. Mid value is an abbreviation for this deterministically computed value.

You can use the Attribute panel in this manner to examine the mid value of any variable in the model.

It is faster to compute a mid (deterministic) value than an uncertain (probabilistic) value, so it is useful for conducting initial checks of a model before performing any uncertainty analysis.

Display results

When you are viewing a model’s influence diagram, you can evaluate any variable and display its value in a Result window.

Chapter 2.27.png

A Result window displays the probability density function graph for this variable. Analytica displays the uncertainty view that was most recently selected from the Uncertainty View popup menu, or that was saved with the model.

Chapter 2.29-updated.png
Tip
The Rent vs. Buy model uses financial flow conventions: funds flowing in (received) have positive values; funds flowing out (expended) have negative values.

As an alternative to clicking the Result button Chapter 2.29.1.png and then selecting an uncertainty view, you can use the Result menu to evaluate a variable and select the uncertainty view of the result.

Chapter 2.30.png

The Result window appears displaying the variable’s cumulative probability distribution.

Chapter 2.31.png
Chapter 2.33.png

A single mid value appears in table view.

Chapter 2.34.png

The mid value in table view is the only result view available for a nonprobabilistic variable with a single value.

Chapter 2.35.png

Analytica tells you that this is a nonprobabilistic value.

Chapter 2.36.png

Summary

You now have browsed the Rent vs. Buy model by examining its influence diagrams, variables, attributes, definitions, and results. These are the basic techniques for exploring any Analytica model.

The next chapter shows you how to analyze the Rent vs. Buy model.

You can quit Analytica at this point.

See Also


Comments


You are not allowed to post comments.