Download


Files
You can download either the 32-bit or 64-bit version of UCINET. The 32-bit version is the standard version and runs on both 32-bit and 64-bit Windows. The 64-bit version is limited in that it does not have all of the functions of the 32-bit version. It often crashes. Therefore, it is best used in tandem with the 32-bit version.

32-bit Installation Package. This installs the 32-bit version of UCINET along with several helper programs (such as NetDraw), and puts a copy of all the standard datasets in a folder called Ucinet Data under your Documents folder. It runs on 64-bit and 32-bit Windows. The installation program is new as of version 6.531 and unfortunately can't automatically uninstall versions of UCINET prior to 6.531, so you should do that yourself prior to installing current version.

64-bit installation package. This is NOT recommended. If you have a 64-bit version of Windows, you can try this 64-bit version of UCINET. It lets you analyze much larger datasets, if you have the RAM memory for it. However, it is also flaky. Older parts of UCINET are not compatible with 64-bit execution, so these parts will crash. 

Trial vs "real" version. Actually, there is no separate trial version. If you download the program, it will run on your computer for 60 days without having to enter a registration code. Buying the program gets you that registration code. So whether you buy or just try, you download the same program.

Virus/malware issues. There was a time when the download files were stored in a DropBox location. For some reason, this caused Norton AntiVirus to flag the program as dangerous. So we don't use DropBox for this anymore. But you might still run into issues with Norton or some other anti-virus program when installing UCINET. One thing to be aware of is that Windows 8 will issue a warning that the publisher is unknown and probably dangerous. This just means we haven't sought (read: paid for) Microsoft certification.


Installation Notes

Windows

Installation works best if you right-click on the installation file (typically called something like setup32UCI6534.exe) and choose 'Run as Administrator'. Similarly, it is also helpful to use 'Run as Administrator' when running UCINET itself for the first time, as this allows you to register the program for all users of your machine. See the FAQ on this.

MAC

The best way to run UCINET on a Mac is to use a Windows emulator such as Parallels (or, of course, Bootcamp). However, it is (often) possible to run UCINET on a Mac using Wine. For more information, see this FAQ

Version Info

  • UCINET 6.620 | 13 June, 2016
    • Fixed important bug in Data|Affiliations|Mode=Rows. When given a multi-matrix dataset as input, it was writing out each 1-mode projection twice.
    • Replaced k-plex routine in Network|Groups|K-Plexes
    Posted Jun 13, 2016, 11:12 AM by Steve Borgatti
  • UCINET Version 6.619 | 10 June 2016
    • Made change in CLI to force use of courier new font. Previously, it was using the default Windows system font, which was almost always courier, but occasionally not. Using courier ensures that columns of numbers line up properly
    [ download
    Posted Jun 10, 2016, 4:51 AM by Steve Borgatti
  • Version 6.618 | 4 June 2016
    • Added new command Data|Filter/extract|Extract Columns. Use it to pull out one or more columns from a dataset and put them in a new dataset.
    Posted Jun 5, 2016, 9:24 AM by Steve Borgatti
  • Version 6.617 | 3 June 2016
    • Fixed bug in Tools|Testing hypotheses|Node level|Regression that was causing the program to always choose the first column of the dependent variable dataset
    • Fixed bug in the matrix editor which was causing crashes on pressing the New button.
    Posted Jun 3, 2016, 3:40 PM by Steve Borgatti
  • Version 6.616 | 30 May 2016
    • Fixed bug in the CLI's Forfiles command. Previously, commands of the form ->forfiles samplk? draw were failing, while commands like ->forfiles samplk* $-cent = centrality($) were succeeding.
    • Note: bug discovered in this version in Tools|Testing hypotheses|Node-level|Regression. No matter which column you select as dependent variable, it always runs the regression on column 1
     download 

    Posted Jun 1, 2016, 5:45 PM by Steve Borgatti
  • Version 6.615 | 29 May 2016
    • Note: bug discovered in this version in Tools|Testing hypotheses|Node-level|Regression. No matter which column you select as dependent variable, it always runs the regression on column 1
    • Added "number of holes" statistic to CLI's Holes routine. It simply counts the number of pair of alters that are not connected. This captures both network size and density
    • Added limited Frequencies command to the CLI. Usage: 
    ->dsp freq(sampson)

    Matrix: Frequencies

              1   2   3   4   5   6   7   8   9  10 
            SAM SAM SAM SAM SAM SAM SAM SAM SAM SAM 
            PLK PLK PLK PDL PES PDE PIN PNI PPR PNP 
              1   2   3   K       S       N       R 
            --- --- --- --- --- --- --- --- --- --- 
        1 0 251 249 250 259 252 248 253 256 267 265 
        2 3  18  18  18  15  17  16  17  16  13  13 
        3 1  18  21  19  15  18  20  17  16  14  14 
        4 2  19  18  19  17  19  22  19  18  12  14 

    Matrix: Proportions

                1     2     3     4     5     6     7     8     9    10 
            SAMPL SAMPL SAMPL SAMPD SAMPE SAMPD SAMPI SAMPN SAMPP SAMPN 
               K1    K2    K3    LK     S    ES     N    IN     R    PR 
            ----- ----- ----- ----- ----- ----- ----- ----- ----- ----- 
        1 0 0.820 0.814 0.817 0.846 0.824 0.810 0.827 0.837 0.873 0.866 
        2 3 0.059 0.059 0.059 0.049 0.056 0.052 0.056 0.052 0.042 0.042 
        3 1 0.059 0.069 0.062 0.049 0.059 0.065 0.056 0.052 0.046 0.046 
        4 2 0.062 0.059 0.062 0.056 0.062 0.072 0.062 0.059 0.039 0.046 


     download 
    Posted Jun 1, 2016, 2:47 PM by Steve Borgatti
  • Version 6.614 | 22 May 2016
    • Note: bug discovered in this version in Tools|Testing hypotheses|Node-level|Regression. No matter which column you select as dependent variable, it always runs the regression on column 1
    • Changed Network|Compare aggregate proximity matrices|partition to be able to handle missing values
    • Changed the CLI's IPF routine to default to treating diagonal values as valid. If you want the diagonal ignored you must add the word IGNORE to the command. E.g.,
      ->dsp ipf(padgm ignore)
     download 

    Posted Jun 1, 2016, 2:48 PM by Steve Borgatti
  • Version 6.613 | 18 May 2016
    • Note: bug discovered in this version in Tools|Testing hypotheses|Node-level|Regression. No matter which column you select as dependent variable, it always runs the regression on column 1
    • Added ability to sort columns in matrix editor via double-clicking column header
    • Added new version of node level regression under tools|testing hypotheses|node level|regression
    • Fixed bug in CLI version of regression which was multiplying classical p-values by df/(df-1). 
    • Fixed recently added bug in tools|testing hypotheses|dyadic|relational cross tabs which was printing incorrect correlation values
     download 

    Posted Jun 1, 2016, 2:50 PM by Steve Borgatti
Showing posts 1 - 8 of 203. View more »

Comments