Executable installer for v. 1.8.3

Installers

This release contains the first version of the all-in-one installer for the app. It allows the user to install the app locally and use all of its functionalities without needing to install R. They are fully update to the complimentary release to version 1.8.3 (the same available also as an R package from the sister project BlockmodelingGUI)

There are two installer available:

• Fully executable (default for most users): Provides a GUI.exe file that can be used to run the app like any other piece of distributed software. This version comes with the warranties indicated in 'Section 5' of the software license (reproduced below);
• Scripted (for advanced user): Provides a run.vbs file that can be used to run the app via Window's PowerShell. This option allows to customise slightly the execution of the app and its behaviour (e.g., by specifying in which of the installed browsers the app should run). Any modification of this sort breaks all warranties (if any would otherwise have been applicable).

Software warranties

To the extent possible, the Licensor offers the Licensed Material as-is and as-available, and makes no representations or warranties of any kind concerning the Licensed Material, whether express, implied, statutory, or other. This includes, without limitation, warranties of title, merchantability, fitness for a particular purpose, non-infringement, absence of latent or other defects, accuracy, or the presence or absence of errors, whether or not known or discoverable.
To the extent possible, in no event will the Licensor be liable to You on any legal theory (including, without limitation, negligence) or otherwise for any direct, special, indirect, incidental, consequential, punitive, exemplary, or other losses, costs, expenses, or damages arising out of this Public License or use of the Licensed Material, even if the Licensor has been advised of the possibility of such losses, costs, expenses, or damages. Where a limitation of liability is not allowed in full or in part, this limitation may not apply to You.
The disclaimer of warranties and limitation of liability provided above shall be interpreted in a manner that, to the extent possible, most closely approximates an absolute disclaimer and waiver of all liability.

v. 1.8.3

Aestetics/Inputs:

• The default interface automatically selected complete and null blocks as default regardless of the approach selected
• The fields to impute the parameters density and average appear or disappear depending on the user's block-type selection thanks to the functions show() and hide() provided in the packaged shinyjs

Server logic and Outputs

• At the block in sect. 2 ('Output'), № 22.1
  • Streamlining calls to UpdateSelectInput()
  • replaced the UpdateBlockTypes() function (a call to observeEvent) with an observer
• Automatic suggested values for the parameters density (network's density) and average (ties' average value, incl. 0s) thanks to the an observer in sect. 2 ('Output'), № 22.2 (Solves #19)
• The user cannot select the option to plot the partitions until the blockmodeling is run. Then, the partitioned network becomes the default option and the user can select the non-partitioned one. (Alternative solution to #20)
• The user cannot select the option to plot the partitioned matrix until the blockmodeling is run. Then, the partitioned matrix becomes the default option and the user can select the non-partitioned one. (Extension of the alternative solution to #20)

Various and eventual

• Corrected some typos in the block's titles
• Fixes: Block № 22 was not appearing in the navigation menu

Complimentary update

Aestetics/Inputs:

• Added an 'Info' section where the user can learn how to credit @aleszib and @FATelarico in papers and other publications for the realisation of which this app was employed.

Server logic and Outputs

• The tab also offer the possibility of downloading all the necessary and recommended references in .ris and .bib formats
• The user can specify to launch browser from command line. This is useful for advanced users who can create a shortcut to launch the app without first launching R (similarly to how the executable version will work, but still requiring to have installed R) [@aleszib]

v. 1.8.2 - Facelift

Main release

Aestetics/Inputs:

• Redesigned aesthetics of the 'Generalised blockmodeling' tab
• Selectors for the number of digits to print for EM and Mean matrix
• Menu to select block-type weights
• Button to restore results
• Counter to set how many results should be stored for restoring (from 1 to 10)
• Selector of plot-matrix edges' size
• Description updated #17

Server logic and Outputs

• Fixed: Mean matrix's and Error matrix's values are not automatically rounded (see #16)
• The block-type weights are taken into account when blockmodeling
• The app stores results and can restore them

Various and eventual

• Corrected comments' levels
• Several sections of code were moved and renumbered

Second update

Aestetics/Inputs:

• The default interface and the one for pre-specified block-models allow different blocks for different approaches
• Selection of parameters for avg and den blocks
• Checkbox to delete isolated nodes

Server logic and Outputs

• Selecting and setting weights for more block-types #18

Third update

Aestetics/Inputs:

• The default interface automatically selected complete and null blocks as default regardless of the approach selected

Server logic and Outputs

• Streamlining calls to UpdateSelectInput() in sect. 2 ('Output'), at the block № 22.1

v. 1.8.1 Rolling updates

Second release

Server logic and Outputs

• Fixed: Error when uploading a bi-dimensional array as a pre-specified block-model

First release

Aestetics/Inputs:

• Plotting with igraph and network :
  • possibility of hiding nodes' labels #14
  • When nodes' labels are hidden, no option related to them can be set
• Added a clarification regarding default arrow settings for plotting with network #13
• Plotting with network, igraph, visNetwork
  • Possibility of selecting the preferred 'best' solution (when more than one exists)
• Selector of the preferred 'best' solution for plotting of the partitioned matrix
• Downloading the clusters to a file now saves the selected 'best' solution for plotting of the partitioned matrix
• Mean matrix as a blockmodeling output
• 'Export adjacency matrix' appears only when visualising the adjacency matrix as a table

Server logic and Outputs

• Fixed: Partioned matrix was not being plotted
• Fixed: Network plots never had arrows by default #13

Various and eventual

• Corrected title of a section in the comments

v. 1.8.0

Aestetics/Inputs:

• Simplified page layout
• Simplified CSS stylesheet
• Redesigned interface for showing adjacency matrix
• Button to download adjacency matrix
• Add attribute is hidden when the 'Sample' network is selected
• visNetwork interface:
  • Select nodes' colour from a list
  • Select edges' highlight colour from a list
  • Select edges' colour from a list
  • Select plot' backgdround colour from a list
  • Node colour's can show attribute
  • Selecte palettes for nodes' colours when using an attribute
• iGraph interface:
  • All colour selections now happen though a list
• Button to download the customised blockmodel
• Panel showing the error matrix

Server logic and Outputs

• Download adjacency matrix (incl. for 'Sample') as a txt file
• Export custom blockmodel after processing the data as a RDS file
• The image-matrix selector affects also which error matrix is seen
• All colours for visNetowrk and iGraph plotting are selected from built-in palettes
• Streamlined the call to visNetwork()
• Minor bug fixes

v. 1.7.4 - Rolling-base enhancements

Aestetics/Inputs:

• Reordered interface for basic blockmodeling
• File upload for network attribute (optional)
• Corrected spelling mistakes and typos
• Improved use of self-disappearing notifications
• Addition of a modal spinner for long blockmodeling operations
• Addition of a loading spinner for other long operations
• Button 'Upload blockmodeling results'
• Button 'Download blockmodeling as file'
• Selector for nodes' colour for igraph when using an attribute do determine the colour
• Selector for clusters' colours for igraph
• Selector for clusters' colours for visNetwork
• Guide to the numbers of colours in each palette
• Warning message for short palettes while using iGraph
• Warning message for short palettes while using visNetwork
• Partioned matrix's plotting:
  • plotting area expanded
  • resolution increased
• Original matrix's plotting
  • plotting area expanded
  • resolution increased
• network and igraph plotting
  • plotting area expanded
  • resolution increased
• visNetwork plotting
  • plotting area expanded
  • increased interactivity

Server logic and Outputs

• Support for busyShiny
• Downloads blockmodeling's results as RDS and reads data from included matrix
• Reads blockmodeling results from an RDS file
• The data from a preloaded blockmodeling result can be reused for a new blockmodeling without having to upload the original data file
• Fix: The blockmodel table is not sortable anymore
• Fix: Trying to plot the partitioned matrix before executing blockmodel will return the original matrix with a warning (see Issue #7)
• Allows the selection of clusters' colouts amongst all available palettes, 'R4' is the default option
• Warning message when the palette does not contain enough colours
• Nodes can be searched by id while plotting with visNetwork
• Correction of a bug happening when a custom block-model contains only one block-type per cluster (see Issue #11)
• Fix: warning on the number of colours triggered when unnecessary
• Add an attribute to the network's vertexes
• Change the colour of the network's vertexes depending on an attribute (igraph)
• Palette selection when polling using an attribute of the network's vertexes to determine their colour (igraph)

v. 1.7.3 - Minor enhancements

Aestetics/Inputs:

• Added a button to 'Select all' cells in personalised blockmodeling
• Added a button to 'Revert selection' of cells in personalised blockmodeling

Server logic and Outputs

• Both buttons leverage the creation of a proxy to manipulate the existing DataTables from the server's side
• The button resetting the selection uses basic JavaScript
• The button selecting all cells uses the function selectCells()

v. 1.7.2 'The Block'

Aestetics/Inputs:

• All labels and selectable options were shortened
• The interface for imputing personalised block-models was simplified and redisigned
• Renaming 'Clusters' size' to 'Number of Clusters'

Server logic and Outputs

• Added support for DT
• Enhanced visual aids (table with cell selection) to build a personalised block-model;
• Bug fixes related to the upload and editing of personalised block-models
• Reduced number of calls to reactive values in the building a personalised block model in more steps thanks to
• The single reacitveEvent for all operations related with personalised block-models was replaced by three separate observeEvent
• Bug fixes in the override of ovveriding manual changes without having to reupload the file containing a personalised block-model

Others and eventual

• Improved code readability and navigability (coherent use of title and sections for the new lines of code)

v. 1.7.0 'The Block'

Aestetics/Inputs:

• Changed blockmodeling approach's selection type from radio buttons to dropdown selection
• Completely new interface for imputing personalised block-models, incl. compatibility with multiple block-types per cell.
• Removed the distinction between 'Cluster rows' size' and 'Cluster columns' size'
• Removed the selectors for the first two dimensions of personalised block-model array
• Ability to select the size of a square block-model matrix
• Selection of the blockmodeling approaches is carried out via drop-down box (instead of radio buttons) to save space.

Server logic and Outputs

• Ability to build a personalised block model using visual aids (table)
• Building a personalised block model using dropdown block-type selection
• Possibility of building a personalised block model in more steps thanks to the use of reactive values
• Ability to upload and edite pre-made block-model arrays with four dimensions
• Possibility to override manual changes without having to reupload the file containing a personalised block-model
• Removed support for HandsOnTableR

Others and eventual

• Improved code readability and navigability (better use of title and sections)

Full Changelog: v1.6.0...v1.7.0

v. 1.6.0 - The Customised

• When extra dimensions are dropped, the user can download the selected image matrix as a text file only or a RDS file
• When extra dimensions are not dropped, the app generates the image matrix only as a RDS file
• The blockmodeling side panel includes the option 'Pre-specify block-model'
• An additional side panel appears when the new box is checked which allows the user to :
  • use a personalised block-model by uploading an array contained in a .RDS or .RData file;
  • use a sample (which corresponds to the block-model used as an example in the package blockmodeling's help file (which is also linked directly);
  • employ a fully interactive table based on handsontable.js. However, this option is only available when the first two dimensions of the array equal 1 and convenient only for smaller block-models;
• The user receives notifications informing of the advancement of the blockmodeling process, which can be quite useful when relatively complex models and/or networks are employed;
• When the user selects to pre-specify block-models, the number of clusters cannot be set (the option is hidden) and it is automatically determined based on the user-provided block-model.