What's fixed in release 4.1.1?

What's new in Analytica 4.1? >>

The 4.1.1 patch release for Analytica and ADE contains over 100 bug fixes and some minor enhancements, including speed improvement of abotu 9%. Because of the bug fixes, we highly recommend that all 4.1 users upgrade to 4.1.1. There is nothing that should impact backward compatibility.

To upgrade to the latest patch release, just download the relevant release (Analytica or ADE 4.1.x) from the Analytica Download page. Run the installer. You will not need a new License Code if you already have 4.1.

A newer patch, release 4.1.2, is available as of 12 Sep 2008. If you upgrade from 4.1.0 to 4.1.2 directly, you will get all these fixes as well. There is no need to install 4.1.1 as an intermediate step.


  • Release 4.1.1 runs approximately 9% faster on average compared to 4.1.0.


  • The Open File dialog now contains a fast links to Analytica Example Models, as well as to other special folders.
  • The shortcut folder in the Start --> Programs --> Analytica 4.1 menu how has a shortcut to the Analytica Wiki. Also, the Help menu in Analytica has a link to the Wiki.
  • Problems scrolling with the mouse wheel were fixed, and support was added for several additional scrolling devices found on various computers), including horizontal scrolling wheels. Also the wheel now scrolls the attribute pane.
  • An output node that evaluates to a handle will now act as a hyperlink to the object.
  • Choice now displays options correctly even when the first option is blank (a zero length text value).

Operating system compatibility

Windows Vista

Several fixes were focused on fixing problems specific to Windows VISTA.

  • Splash screen not staying over application window
  • Non-functional close buttons on some dialogs

Windows 2000-specific

  • Several incompatibilities with Windows 2000 were discovered, dependent on how much RAM the computer had. These have been fixed so it runs gracefully once again in Win 2000.


  • Small change to color sequence so the 5th color is more visible.
  • Bar overlap attribute didn't "Set Default" correctly.
  • CDF graphs now always start at zero. Some graphs used to have a gap at the left when the first CDF value was non-zero.


  • A serious bug existed with Sum, Max, and other related functions that was fixed. In some rare cases, when used with multiple or nested indexes, with some of the indexes being sparse (an internal distinction that can arise when values are constant along certain slices), an incorrect answer was being returned. This is a primary reason why we urge 4.1.0 users to upgrade, even though this problem shows up very rarely.
  • Subscript(A,I,1) or A[I=1], would fail to perform the subscript in some rare cases, returning the entire array.
  • System function Min wasn't returning NaN in some cases where the array contained NaN values and ignoreNaN was false.
  • Today(withTime:true) was off by one hour during daylight savings time.
  • Function Combinations(0,N) now returns 1, rather than an error message.
  • Text values were previously limited to 50M characters. Exceeding this limit could result in a crash. The limit has been relaxed (to available memory, up to 4 billion characters).
  • Index parameters are now optional to Function MatrixMultiply, so that it can be used to multiply matrices and vectors, for example.
  • The howToIndex option to use the first row in WorksheetRange now works correctly.
  • Round now rounds correctly when numbers have more than 12 digits.
  • When a value containing warnings is displayed using MsgBox, the value and warning text are now shown, to help during model debugging.
  • Slice(A,i) now allows both A and i to be unindexed lists, without an error message. Also, Slice(A,i) previous did not allow A to be an unindexed scalar, but now does.
  • HandleFromIdentifier was returning Undef, rather than Null, when the object didn't exist. It now returns Null. To be forward compatible with this check, to test whether an object exists use IsUndef(HandleFromIdentifier(v)) and it will also work in 4.0 and 4.1.0.
  • The date origin when using the "Use Excel Date Origin" preference now gives results consistent with Excel.
  • A memory leak with the undo edit buffer was fixed, which occurred when a choice input was changed and results were recomputed.
  • The date/time parser (parsing a cell or definition when number format is set to date) now recognizes additional date-time formats, such as yyyy-mm-dd hh:mm:ss, or "9:40a" for "9:40 am".

OLE Linking

  • OLE links out of Analytica were previously sent in the number format that was active at the time the link was established. So if you were viewing the result table with 4 digits, the links would go out with 4 digits. OLE links out of Analytica are now always full precision.
  • When viewing an OLE linked edit table, in edit mode, the right-mouse menu now contains a "Break Link" option. Previously you had to do this from the Edit --> OLE Links... dialog, where it was often hard to tell which link was which.


  • The wrong number format setting was used for some Pdf columns.
  • With some very complex dependencies between computed indexes and spliced tables, involving many dependency loops, some tables would not splice successfully when computed indexes changed.
  • Check attribute for cell-by-cell checking was made a bit more flexible.


  • The Check value bounds preference didn't suppress checking of UDF parameters or of cell-by-cell checks in edit tables. Now it does.


  • Some efficiency shortcuts were added to the typescript window.
    • CTRL-Up and CTRL-Down cycle through a history of commands previously typed in the command window, so you can make minor editing changes without retyping the entire line.
    • Tab auto-completes identifiers. Pressing it twice shows all possible completions. Without the shift key, limits completions to user symbols. Shift-tab includes all symbols, user and system, in the completion.
    • You can continue typing the same command on a second line if the last character is a tilde (~)

Example Models and Libraries

  • ReadCsvFile was ignoring the optional colSep and lineSep parameters.

User Guide

  • The User Guide and Tutorial now open at 100% size in Adobe Reader.
  • We have fixed several bugs and typos, along with a few new index entries.
  • We added more on using the Check attribute, especially for cell-by-cell checks.


  • When ResultType=1 (mean), and a result table contains textual values that don't vary over Run or are not probabilistic, the text values are now returned rather than NaN.
  • Several reported bugs that resulted in crashes or exceptions were fixed.

You are not allowed to post comments.