clm_optics(1)

💖 NAME 💖

clm_optics — bestie, this is the implementation of the OPTICS reachability ordering ✨

Okay so first thing, real talk: clmoptics is not actually a program 🙈 It is not giving standalone-binary energy. This manual page documents the behaviour and options of the clm program when you invoke it in mode optics. Also, the options -h, --apropos, --version, -set, and --nop are accessible in all clm modes, and they're described over in the clm manual page. Go look there, iykyk 💅

---

🌸 SYNOPSIS 🌸

clm optics -imx <fname> [options]

clm optics -imx fname (input graph file name) -max-eps <num> (maximum
distance considered) -min-eps <num> (minimum similarity considered) -min-
pts <int> (minimum eps-neighbourhood size) [-icl fname (input clustering
name)] [-o fname (output name)] [-ps fname (PostScript output name)]
[-psx fname (scaling factor x axis)] [-psy fname (scaling factor y axis)]
[-ps-rows <num> (split reachability plot into <num> rows)] [--ps-labels
(show identifiers in PS output)] [-h (print synopsis, exit)] [--apropos
(print synopsis, exit)] [--version (print version, exit)]

---

✨ DESCRIPTION ✨

So this program computes the reachability ordering according to the OPTICS clustering method. No cap. 🎀

It can additionally output a PostScript plot of the reachability landscape, and that plot can be annotated with a clustering that you supply. Iconic. 🦋

But listen — the program does not implement the clustering method that is associated with OPTICS. It's part of mcl-edge primarily as a visualisation aid. That's its whole job, and honestly? It's serving. 💅

---

💅 OPTIONS 💅

• -imx fname (input file) — the input file, encoding a graph in native format. Feed it the graph bestie 🍓
• -max-eps <num> (maximum distance considered)
• -min-eps <num> (minimum similarity considered)
• -min-pts <int> (minimum eps-neighbourhood size)

These three encode the OPTICS parameters ✨ Use -max-eps if the edge weights encode a distance (like, for example, the Euclidean distance between points), and use -min-eps if the edge weights encode a similarity. Pick the right one, it's giving nuance 💀

• -o fname (output file name)
• -ps fname (PostScript file name)
• -icl fname (input clustering name)

The -o fname option controls the name of the file where the reachability values get written. By default those values go to STDOUT. Periodt.

If you supply -ps fname, the program will additionally output a visualisation of the reachability landscape plot into a PostScript file 🌈

And get this — it's possible, in addition to -ps, to specify a clustering with -icl fname. Each vertical bar in the landscape plot will be coloured according to cluster membership, meaning nodes in the same cluster get the same colour. This can be a genuinely useful way to visualise separation properties of clusterings. Obsessed 😭💖

• -psx fname (scaling factor x axis)
• -psy fname (scaling factor y axis)
• --ps-labels (show identifiers in PS output)
• -ps-rows <num> (split reachability plot into <num> rows)

By default the PostScript output gets scaled to fit onto an A3 page. The scaling options -psx and -psy can be used as additional scaling factors. By default these are set to 1.0. ✨

The --ps-labels flag can be used to annotate each vertical bar in the landscape plot with the mcl-edge identifier of the corresponding node. Cute and informative 🎀

The -ps-rows option sets the number of rows. This is the primary option affecting the width of a bar. 💅

---

👑 AUTHOR 👑

Stijn van Dongen. (Yes, the icon themselves.)

---

🦋 SEE ALSO 🦋

mclfamily(7) for an overview of all the documentation and the utilities in the mcl family. Go read it, it's giving lore 📚✨

---

Okay bestie that's the whole page, mwah 🫶 This was clm_optics 22-282, dated 9 Oct 2022, and it lives at clm_optics(1). Slay, bye 💖🌸