HumdrumR

[Benjamin Bradshaw]

Hello all,

I'm still at the point of wanting to learn humdrum (by going through the user guide). I know there's a new HumdrumR. Is it best to avoid this for now (especially since I don't know R either) and focus on humdrum basics? Or does HumdrumR bring enough features that I should start with it? Also, are there other great sources for learning how to do this kind of computational musicology?

Thanks,

Ben

[Eleanor Selfridge-Field]

Hi, Ben.

We have Humdrum tutorial material online at our teaching websites—music253.stanford.edu (Symbolic Musical Information) .  See Week 9 of the annotated syllabus.  

https://wiki.ccarh.org/wiki/Music_253/CS_275a_Winter_2025_Syllabus

Eleanor

Eleanor Selfridge-Field

Adjunct Professor

Stanford University

Stanford, CA 94305-3076

esf...@stanford.edu

ORCID ID 0000-0003-4618-1372

---

From: 'Benjamin Bradshaw' via starstarhug <stars...@googlegroups.com>
Sent: Wednesday, September 10, 2025 2:47 PM
To: starstarhug <stars...@googlegroups.com>
Subject: [**HUG] HumdrumR

 

--
--
This is a message is from the **HUG newgroup.
To post to this group, send email to stars...@googlegroups.com
To unsubscribe from this group, send email to
starstarhug...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/starstarhug?hl=en
---
You received this message because you are subscribed to the Google Groups "starstarhug" group.
To unsubscribe from this group and stop receiving emails from it, send an email to starstarhug...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/starstarhug/d0701cae-cb03-4511-bf35-c52128809d64n%40googlegroups.com.

[Yuri Broze]

Hi Ben,

To echo Prof. Selfridge-Field, if you'd like to use the Humdrum Kern specification, it's best to get off the ground using "Humdrum Classic" as it were. HumdrumR is a library to help interface the R statistical programming language with Humdrum, to make it easier to do statistical analyses.

Humdrum itself is a collection of bash scripts, awk scripts, and a few C libraries. Very simple, straightforward, hence, "humdrum." Very user friendly and intuitive. Nat Condit-Schultz and Claire Arthur, my former colleagues in David Huron's group at Ohio State, are developing HumdrumR at Georgia Tech. But it's newer, not as "tried and true" just yet.

I'd follow the CCARH guide, since Craig Sapp has been developing and maintaining the core functionalities while Unix like distributions spin around like a top.

Yuri

Yuri Broze
314.910.3152

To view this discussion visit https://groups.google.com/d/msgid/starstarhug/SJ0PR02MB8419E3AB12C29285AEB43B41C30EA%40SJ0PR02MB8419.namprd02.prod.outlook.com.

[Joshua Veltman]

Fun fact: when I was a student of David Huron's some 25 years ago (wow, how did that happen?), he mentioned to me that he had considered calling the software "hohum" because he didn't want to bother with patenting / copyrighting it and wanted to pick a "boring" name that no company would ever want to use for a product. He ended up picking "humdrum" for the same reason, but also because that name fits better with the musical purpose, as "hum" refers to pitches and "drum" refers to rhythms. 🙂

To view this discussion visit https://groups.google.com/d/msgid/starstarhug/CAKxiJpUArixiW91_sLVG6dgE_Qm-aB%2BouG1b2XX0WTcFRE5Uaw%40mail.gmail.com.

[Nathaniel Condit-Schultz]

As the main developer of humdrumR, I can address Ben's question from my perspective, but this is sort of addressed at the whole group. I think the ideal relationship between humdrum and humdrumR has yet to be really worked out.

We are still working out for ourselves whether, pedagogically, it makes sense for users to first learn the original humdrum toolkit (and humdrum extras) before learning humdrumR. In some ways it makes sense to do so, and humdrumR actually emulates the original toolkit in many ways, so it can work. However, not a lot of people are interested in learning (or installing) bash/awk based tools at all; Learning R/RStudio is more appealing, especially since you eventually need something like R (or Python) to do stats, plots, or more sophisticated analyses. It can feel (for students) like, "why are we learning these old-fashioned bash tools if we won't be able to do the full analysis with them anyway"? (I use bash all the time myself.)

Ultimately, I do envision that new humdrum users could indeed "skip" the classic humdrum tools and start with humdrumR. humdrumR can (or will eventually be able to) do everything the humdrum/extra toolkits can do, and it also enables a lot of things that they can't, including but not limited to statistics. That being said, the things that "classic"/extra tools can do, they can do faster and cleaner than humdrumR. For example, the mint command is amazing. You can do it in humdrumR, but the ease and speed of the original mint crazy.  

Here is a scenario: If you want to find the most common scale degree in the soprano voice of the Bach chorales, you can do that faster and (slightly) more simply in the original humdrum toolkit, compared to humdrumR. But if you want to find the most common scale degree in each of the four voices, it will probably be just as easy to use humdrumR. And, if you want to find the most common scale degree used in each voice, but only in the upper half of each voice's vocal range and only for durations longer than an eighth note, it's going to be much easier to do that in humdrumR. 

Of course, as Yuri pointed out, humdrumR is still in development, and still has bugs and features not yet developed/implemented/documented. I can tell you that I use bash a lot personally, just for day to day work, and of course I use humdrum data almost exclusively in my work, but I don't think I've even had the humdrum tools installed on my machine for years. I use humdrumR exclusively...though to be fair, I am the only person who entirely understands how it works.

Nat Condit-Schultz

Computational and Cognitive Musicology Lab

School of Music, Georgia Institute of Technology

To view this discussion visit https://groups.google.com/d/msgid/starstarhug/CAM2xP86eaz3%3DQzPVkYxXfXCXLUVB9RbLxMqdHnRCFhDjEuiPOQ%40mail.gmail.com.

[Yuri Broze]

Joshua,

This fact is, indeed, quite fun!

Yuri

Yuri Broze
314.910.3152

To view this discussion visit https://groups.google.com/d/msgid/starstarhug/CAM2xP86eaz3%3DQzPVkYxXfXCXLUVB9RbLxMqdHnRCFhDjEuiPOQ%40mail.gmail.com.

[Benjamin Bradshaw]

Thanks for the resources and advice all! I think I'll continue with the classic humdrum for now. I'm at least comfortable with bash/awk and have no experience with R at all! So it seems like an easier introduction. But that is an interesting question for the future. I could see if someone wasn't familiar with bash/awk or unix stuff in general that maybe humdrumr could be less new stuff to learn.

Best,

--

BENJAMIN BRADSHAW, DM   |   ADJUNCT PROFESSOR MUSIC DEPARTMENT SOUTHERN UTAH UNIVERSITY Center for Music Technology

 

On Sat, Sep 13, 2025 at 5:55 PM Yuri Broze <yuri...@gmail.com> wrote:

Joshua, This fact is, indeed, quite fun! Yuri Yuri Broze 314. 910. 3152 On Fri, Sep 12, 2025 at 4: 32 PM Joshua Veltman <joshua. veltman@ gmail. com> wrote: Fun fact: when I was a student of David Huron's some 25 years ago (wow, how did

To view this discussion visit https://groups.google.com/d/msgid/starstarhug/CAKxiJpVmB-z5SjCER3D0oUPpgxcHVt%3D%2BPUUQBi%3DJsf20DSgMjQ%40mail.gmail.com.

[Nathaniel Condit-Schultz]

Huron told me the same explanation about why he chose "humdrum." But I don't think he mentioned that "hohum" was on the table. Holy toots! That's hilarious.

Nat Condit-Schultz

Computational and Cognitive Musicology Lab

School of Music, Georgia Institute of Technology

To view this discussion visit https://groups.google.com/d/msgid/starstarhug/CAM2xP86eaz3%3DQzPVkYxXfXCXLUVB9RbLxMqdHnRCFhDjEuiPOQ%40mail.gmail.com.