Versions

This page shows changes to UCINET between versions. Just scroll down to see release notes in reverse chronological order.
Note: To find revision information for really old versions of UCINET look in these two places:
Related Links
  • To see a list of known bugs and user wish list items, check the fix list
  • To download the latest version, click download

News

  • Week-long workshop on SNA The LINKS Center at the University of Kentucky is offering its annual 1-week summer workshop on social network analysis June 6-10, 2016 on the University of Kentucky campus ...
    Posted Mar 15, 2016, 12:54 PM by Steve Borgatti
Showing posts 1 - 1 of 9. View more »

UCINET 6.620 | 13 June, 2016

posted Jun 13, 2016, 11:11 AM by Steve Borgatti   [ updated Jun 13, 2016, 11:12 AM ]

  • 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

UCINET Version 6.619 | 10 June 2016

posted Jun 10, 2016, 4:49 AM by Steve Borgatti   [ updated Jun 10, 2016, 4:51 AM ]

  • 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

Version 6.618 | 4 June 2016

posted Jun 5, 2016, 9:23 AM by Steve Borgatti   [ updated Jun 5, 2016, 9:24 AM ]

  • 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.

Version 6.617 | 3 June 2016

posted Jun 3, 2016, 3:40 PM by Steve Borgatti

  • 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.

Version 6.616 | 30 May 2016

posted May 30, 2016, 3:14 PM by Steve Borgatti   [ updated Jun 1, 2016, 5:45 PM ]

  • 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 

Version 6.615 | 29 May 2016

posted May 30, 2016, 7:07 AM by Steve Borgatti   [ updated Jun 1, 2016, 2:47 PM ]

  • 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 

Version 6.614 | 22 May 2016

posted May 22, 2016, 11:02 AM by Steve Borgatti   [ updated Jun 1, 2016, 2:48 PM ]

  • 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 

Version 6.613 | 18 May 2016

posted May 18, 2016, 7:35 PM by Steve Borgatti   [ updated Jun 1, 2016, 2:50 PM ]

  • 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 

Version 6.611 | 8 May 2016

posted May 8, 2016, 1:21 PM by Steve Borgatti   [ updated May 8, 2016, 1:22 PM ]

  • Adds countcombinations routine to the CLI. Given a matrix of 1s and 0s, it counts up the number of combinations of column items found across all rows of the matrix. For example, if applied to the 2-mode Davis data, it would tell you how often each different combination of events occurs in the data. Syntax is CountComb(<dataset>). Example:

    ->dsp countcomb(davis)
Frequencies
                                         1     2 
                                      Freq  Prop 
                                     ----- ----- 
    1     E2, E3, E4, E5, E6, E8, E9     1 0.056 
    2         E1, E3, E5, E6, E7, E8     2 0.111 
    3     E2, E4, E5, E6, E7, E8, E9     1 0.056 
    4                     E3, E4, E7     1 0.056 
    5                     E3, E5, E8     1 0.056 
    6                     E5, E6, E8     1 0.056 
    7                         E6, E9     1 0.056 
    8                     E5, E7, E8     1 0.056 
    9                E7, E8, E9, E12     1 0.056 
   10               E8, E9, E10, E12     1 0.056 
   11          E8, E9, E10, E13, E14     1 0.056 
   12      E7, E8, E9, E10, E12, E14     1 0.056 
   13 E6, E7, E9, E10, E11, E12, E13     1 0.056 
   14          E7, E8, E10, E11, E12     1 0.056 
   15                         E8, E9     1 0.056 
   16                        E9, E11     2 0.111 

In these data, only two combinations of events occur more than once. 


Version 6.610 | 25 April 2016

posted Apr 25, 2016, 7:34 PM by Steve Borgatti   [ updated Jun 1, 2016, 2:49 PM ]

  • 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
  • Fixed bug in matrix editor that was causing it to save matrices with more rows and columns than were actually present
  • (Also, a bug was fixed a few versions ago that was causing "argument out of range" errors, but the fix was not previously reported)
Update 7 May 
  • Updated the installer to include latest version of NetDraw (2.159), which fixes a bug in reading 2-mode data

 download 

1-10 of 203