# Pecuzal Tmax

**URL:** <https://discourse.julialang.org/t/pecuzal-tmax/111436>\
**Category:** General Usage\
**Created:** [March 10, 2024, 11:28pm UTC](https://discourse.julialang.org/t/pecuzal-tmax/111436 "2024-03-10T23:28:35Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![ccccc](https://avatars.discourse-cdn.com/v4/letter/c/aca169/32.png) [@ccccc](https://discourse.julialang.org/u/ccccc)\
**Post date:** [March 10, 2024, 11:28pm UTC](https://discourse.julialang.org/t/pecuzal-tmax/111436/1 "2024-03-10T23:28:35Z")

</div>

Does anyone know what’s the relation of the PECUZAL algorithm to Tmax? What’s a reasonable Tmax to use?  
I had thought that length(dataset) ÷ 2 - 1 should be ok, but I get a bounds error when I go up this high (only, sometimes, not always). I also noticed that the result can change slightly when I change Tmax, even when both suggested embeddings are well under Tmax, so I wondered if there is a heuristic for the maximum in relation to the length of the dataset?