NewsConsultingCompanyContact Us
What's new in Analytica 4.1?

What is Analytica?
Free trial of Analytica
Order latest release

Here are highlights of new and improved features in Analytica 4.1 added since the release 4.0. Additional detail can be found on the Analytica Wiki

Order upgrade to 4.1,
User Guide A nice introduction to Intelligent Arrays has been added to Chapter 11, "Arrays and Indexes". Intelligent Arrays and intelligent array abstraction are seen by many users as the most useful features of Analytica, and we hope this new introduction will help you gain mastery of this powerful functionality.
 
Installer Analytica 4.1 has an all-new installer. It provides the option of cleanly uninstalling previous releases while upgrading, provides cleaner upgrades and uninstalls, and runs much faster.
 
User Interface  
Images

When you copy/paste a bitmap image into a diagram, the image is automatically compressed into a PNG format to reduce memory and model file size. The new Change Picture Format dialog, accessible from the Diagram menu, allows you to convert the internal format to other available formats, including legacy bitmap if it needs to be viewed in Analytica 4.0 or earlier.

32-bit bitmap images, and images with transparency and alpha blending, are now supported.

Bitmap images draw much more quickly.
Diagrams Diagrams can be exported directly to an image file in PNG, EMF, JPEG, BMP, or TIF formats.

The Set Diagram Size option on the Diagram menu has been eliminated. It was never very useful.
Graphs and charts When you click a data point, a balloon shows the full coordinates of that point at full precision.

Autoscaling is improved when the other axis is manually scaled.
Time formats Using a custom format, the time of day can now be included as part of a date number format. The fractional part of an integer indicates the time as a fraction of a full 24-hour day.
Make importance option The Make Importance option creates two nodes: one to hold the uncertain inputs and one to hold the importance result. Both are now variable nodes, while previously an index node was used for the inputs. The change was made because an index should contain a one-dimensional result, which is not the case for the inputs. Hence, a variable node is more consistent.
Windows Vista A quick animation when windows are opened has been removed. This animation usually lasts for about 1/10 of a second, but in the Vista Aero color scheme was taking over 10 seconds.
File saving File Open and Save dialogs are now resizable.

If you haven’t set the title of your model, it is set automatically to the file name when saved.
Excel Integration  
Copy/paste Cell ranges containing multi-line cells (where one cell contains multiple lines of text) can now be copied and pasted in both directions.
OLE Linking You can OLE link a named range of cells in an Excel spreadsheet into Analytica. The advantage of this is that the Analytica model can automatically adapt when the number of rows or columns in the named range changes without having to adjust the link itself. OLE linking from a named range to any other application, even within Microsoft Office itself, doesn’t work in general due to a bug in Excel (at least in Excel 2003), but Analytica now successfully works around that bug.
Reading Excel files The Enterprise Edition now includes three new functions that read directly from an Excel workbook file: OpenExcelFile, WorksheetCell, and WorksheetRange. These are still considered experimental in release 4.1, so they don’t yet appear on the database Definition menu and are not documented in this guide. Consult the Analytica Wiki for documentation on these functions
Expressions  
New Functions TextTrim(t) removes leading or trailing spaces.

MakeTime(h,m,s) returns the fraction of a day represented by a given time.

ComputedBy(X) indicates that the value of a variable is computed as a side-effect during the evaluation of another variable, X.

OpenExcelFile, WorksheetCell, and WorksheetRange read values directly from an Excel workbook file. These require Analytica Enterprise and are documented on the Analytica Wiki.

IsResultComputed(X) tests whether the result for X is already cached.

BesselJ(x,n), BesselY(x,n), BesselI(x,n), and BesselK(x,n) compute Bessel functions.

Enhancements to Existing Functions Functions Sum, Min, Max, Average, and Product now allow multiple indexes to be specified in a single call when reducing across multiple indexes. For example, Sum(X,I,J,K).

Functions Sum , Min , Max, Argmin , Argmax , Average , Product , JoinText, Irr , and Npv now ignore any null-valued cells in the array parameter, performing their respective operation only over the non-null values. Regression ignores any data points having a null Y value. There is also a new parameter data-type qualifier, OrNull.

Functions Sum, Min, and Max now accept an optional ignoreNonNumbers parameter. For example, Min(X,I,IgnoreNonNumbers:True) finds the minimum among the numeric values of X, ignoring textual or other values.

Functions Sum, Min, Max, Average, and Product accept an optional ignoreNaN parameter.

Function Round accepts a second optional parameter, Round(x,digits).

DatePart has been extended with several new options for accessing the time fields (hours, minutes, and seconds), and the number of elapsed weeks or week days from either the date origin or the beginning of the year. The time field offsets (hours, minutes, seconds) are accepted by the DateAdd function.

The Today function can optionally return the current time as well as date, and also optionally return the date/time in coordinated universal time (utc) rather than in the local time zone.

The Irr function function finds a solution more reliably when a solution exists.

StudentT now supports Latin hypercube sampling.

Shuffle accepts an optional Over parameter (for independent random shuffles).

Concat accepts scalars for its first two parameters when its index is specified. Also, the result index, K, is now optional.

When x is array-valued, the result of Elasticity(y,x) is now different, being now more consistent with the definition of Elasticity being the percent change in y when x changes by 1%.

DbTable now returns Null, rather than NaN, when a numeric value is missing in the relational table.

When Random is evaluated in a sample context, the parameters of the distribution are now also evaluated in sample context.
Meta-indexes Suppose you want a global index to contain a list of identifiers. Analytica’s default evaluation rules with evaluate those variables when the value of the index is requested. In some cases, you might really want just a list of handles (i.e., the index is a "meta-index"). The metaOnly attribute can be set to 1 on a global index object that is defined as a list of identifiers to force this behavior.
For more...

For more details, dowload the Analytica 4.1 User Guide from the Lumina Download Center. For earlier releases, see What's new in Analytica 4.0.
 
Copyright 2008, Lumina Decision Systems, Inc.