DSM Optimizer

Input

Drop a file, or click to browse

.xlsx / .xlsm / .csv matrix · edge-list .csv · .dsmproj project

·

Project

A project saves your entire session - matrix, all analysis results, what-if edits, stability runs, and settings. Save one per design revision, then Compare revisions later to track exactly which dependencies changed and how changes propagate.

Matrix type & convention

DSM type

Groups elements into modules that minimize cross-module dependencies.

Mark convention

How your file's marks read. Wrong choice silently flips what "feedback" means - check one known dependency after loading. Exports use the same convention.

Clustering constraints

The k range guides Spectral & MCL candidates. Cluster-size limits are enforced when you sequence your chosen candidate - raw candidates may sit outside them.

Min clusters (k)

Max clusters (k)

Min cluster size

Max cluster size

Advanced

Bus threshold (connectivity fraction)

Leave blank for adaptive detection (recommended).

Target max external coupling

Advisory only - flags the result if exceeded.

Random seed

Set a value for identical output across runs.

Compare

Load a second DSM to diff…

Highlights added / removed / changed dependencies and computes change-propagation risk on the current matrix.

You'll compare four algorithms and pick one to sequence.

No matrix loaded yet

Load a DSM file, an edge-list CSV, a saved project - or start a blank matrix and build it in the editor. Pick the matrix type (component vs. process) and run.