Office Hours (2026-01-06)

[kgryte]

🌟 Join Us for Weekly Office Hours! 🌟

Have questions about stdlib? Need help fixing a bug, figuring out what to do next, or just looking for feedback? Join our weekly office hours to connect with project maintainers, stay updated on the latest project news, and chat with other community members. This is a great opportunity to ask questions, share ideas, and engage directly with the stdlib team.

Everyone is welcome—drop in and say hello!

Time

UTC Tue 06-Jan-2026 17:30 (05:30 PM):

Timezone	Date/Time
US / Pacific	Tue 06-Jan-2026 09:30 (09:30 AM)
US / Mountain	Tue 06-Jan-2026 10:30 (10:30 AM)
US / Central	Tue 06-Jan-2026 11:30 (11:30 AM)
US / Eastern	Tue 06-Jan-2026 12:30 (12:30 PM)
EU / Western	Tue 06-Jan-2026 17:30 (05:30 PM)
EU / Central	Tue 06-Jan-2026 18:30 (06:30 PM)
EU / Eastern	Tue 06-Jan-2026 19:30 (07:30 PM)
Moscow	Tue 06-Jan-2026 20:30 (08:30 PM)
Chennai	Tue 06-Jan-2026 23:00 (11:00 PM)
Hangzhou	Wed 07-Jan-2026 01:30 (01:30 AM)
Tokyo	Wed 07-Jan-2026 02:30 (02:30 AM)
Sydney	Wed 07-Jan-2026 04:30 (04:30 AM)

or in your local time:

• https://www.timeanddate.com/worldclock/fixedtime.html?msg=stdlib%20Office+Hours+2026-01-06&iso=20260106T1730

Links

• Public calendar: https://calendar.google.com/calendar/embed?src=a72677fe2820c833714b8b9a2aa87393f742bcaf0d0f6c9499eee6661795eae0%40group.calendar.google.com

Joining the meeting

• Google Meet link: https://meet.google.com/yct-chmu-owx

Agenda

• Have specific things to discuss? Please comment in the issue thread below to add specific items to the agenda.

[manit2004]

💬 Items for discussion

• Can @kgryte , @Planeshifter comment on PR #9502 #9324 ?
• I opened issue #9567 for discussion.
• I need to discuss about the sheet I made about all the statistical distributions present in other libraries like R, Julia and SciPy along with feasibility of implementation in stdlib. I have also made a topic for discussion in Zulip stats channel for the same.
• Also I would like to know if I aim to target this gsoc idea of implementing broader range of statistical distributions how should go about it and any guidelines on writing the proposal?
• Create an issue as discussed in the previous meet to replace delta tolerance tests with ulpdiff in all the test packages.
• Should we include variance and mean of the random variable 1/X as well in the wald distribution. Ref

[nakul-krishnakumar]

I would like to discuss about:

• Refactoring of @stdlib/ml/incr/kmeans to include new implementation of @stdlib/stats/strided/distances packages.
• Possibility of adding a general purpose function like scipy.spatial.distance.cdist to replace this.
• Possibility of adding pairwise distance metrics (https://scikit-learn.org/stable/modules/generated/sklearn.metrics.pairwise_distances.html).
• Possibility of adding C implementation for @stdlib/ml/incr/kmeans.
• Possibility of implementing a ml/utils namespace to include utility packages that are going to be common across various ml packages that are to be added. These might include various learningRate decay, weight initialization and other algorithms.
• Implementing packages to finish off these TODOS:
  1. copy_matrix.js
  2. copy_vector.js

@kgryte

[Amansingh0807]

Hi all,
I wanna discuss about Architecture decisions for Issue #44 StringArray support

I'm exploring the StringArray implementation and need clarity on a few design decisions before proceeding:

Critical questions:

1. Memory Layout - Which approach should stdlib use?

   • Offset buffer + data buffer (like Apache Arrow/NumPy NEP-0055)
   • Length-prefixed strings in single buffer
   • Other approach?

2. Element Mutation - When set(index, newString) has a larger string:

   • Reallocate entire buffer (flexible but O(n))
   • Fixed capacity per element (fast but limited)
   • What's the stdlib philosophy here?

3. Backing Store - Preferences?

   • Uint8Array (browser-compatible)
   • Node.js Buffer (better C interop)

4. Worth pursuing now? - Is this a good focus for contributions, or should I prioritize other areas?

5. Resources - Beyond NumPy NEP-0055, what should I study?

   • Apache Arrow strings?
   • Existing stdlib code patterns?
   • Other references?

@kgryte @Planeshifter

[Tejasvini595]

Hi all, I wanted to discuss the following points

1. I raised a PR last week on issue number #730, PR #9450, addressing the creation of erfcf function. Comments were given by @kgryte which I fixed but it hasnt been merged or the labels hasn't been changed from needs changes. Just wanted to catch up on the status of that.
2. Are there any similar issues I can work on related to the above issue, as I already have a hold on how to create such functions, and wish to apply it to something solid.
3. I was interested in this idea I saw on the gsoc idea page on Functions for numerical integration and differentiation, I wanted to know if this idea is still relevant, if I could start working on it, and could it be a possible work for gsoc as well?

[navyansh007]

I'm working on PR #8187 (C implementation of sici). The initial JS test tolerances were too strict for the C implementation (ULP errors of 1-4).

I've implemented tiered tolerances using max(absoluteTol, relativeTol * abs(expected)) with multipliers from 10× to 1000× machine epsilon based on value magnitude.

Questions:

1. Is this the correct testing strategy for C implementations?
2. Should I use @stdlib/assert/is-almost-same-value instead?

[kgryte]

Thanks everyone!