techhub.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
A hub primarily for passionate technologists, but everyone is welcome

Administered by:

Server stats:

4.6K
active users

#multivariate

0 posts0 participants0 posts today

Tile and glyph map experiments: David O’Sullivan explores innovative visualizations of complex #multivariate data on #maps, using a case study with three variables. His article showcases creative #mapping techniques pushing the boundaries of traditional GIS #visualization.
spatialists.ch/posts/2025/05-0 #GIS #GISchat #geospatial #SwissGIS

Spatialists – geospatial newsTile and glyph map experiments – Spatialists – geospatial news
More from Spatialists

Día 15: Complicated Relationships! 🐧↔️🐧

Hoy, una matriz de scatter plots con ggpairs para explorar las relaciones entre medidas corporales (Long. Pico, Long. Aleta, Masa Corporal) en los pingüinos de Palmer. ¡Perfecto para el prompt "Complicated"!

La matriz lo enseña todo:
* Diagonal: Distribución de cada medida (densidad).
* Abajo: Scatter plots de cada par de medidas (coloreado por Especie).
* Arriba: ¡La correlación $ entre ellas!

Se ven las fuertes relaciones positivas (más grande = aleta más larga) y cómo las especies (Adelie, Chinstrap, Gentoo) forman clusters distintos en este espacio de rasgos. ¡Una forma densa de ver muchas relaciones a la vez!

🛠 | Data: | Theme:
📂 Código/Viz: t.ly/GATJi

Anyone know of software that will help optimize multi-variate experiments? Something sample-efficient since these are manually generated test results.

I found github.com/yunshengtian/AutoOE which looks great, but having a hard time getting it working at the moment

GitHubGitHub - yunshengtian/AutoOED: AutoOED: Automated Optimal Experimental Design PlatformAutoOED: Automated Optimal Experimental Design Platform - GitHub - yunshengtian/AutoOED: AutoOED: Automated Optimal Experimental Design Platform

Our article @MethodsEcolEvol on incorporating predictors in a model-based ordination is now out!

We develop a new implementation for constrained ordination in the #GLLVM R-package, but also develop a new kind of ordination.

Concurrent ordination combines the best of unconstrained and constrained ordination. It leaves the latent variables unconstrained but uses predictors to inform the ordination anyway!

besjournals.onlinelibrary.wile