Robust, efficient equation numbering in MS Word 2016

April 29, 2018 6:02 PM

I'm writing a doctoral thesis in mechanical engineering. I will have lots of equations to write in my paper which need to be labeled with numbers. Throughout the body of the paper I will have to cross-reference those equation numbers in discussion. I am currently using the built in functionality in Word but it is not tailored for equation numbering and includes some very cumbersome and finicky workarounds. How do people handle this efficiently in their science/engineering papers? Are there 3rd party products that will do this for me automatically and update the numbers automatically for both the equations and cross-references to them? I am willing to spend money on a solution if need be. The solution must work with MS Word 2016 and I also need the ability work on the same document on two different computers' installs of Word 2016. Any help is appreciated.
posted by incolorinred to Computers & Internet (17 answers total)

Best answer: Have you tried LaTeX? Word is godawful for what you're trying to do, but I guess you can export from pdf to .doc and have it work? There is quite a learning curve to LaTeX but you're a mechanical engineer, and I assume you're at least somewhat acquainted with programming because of that, so it should not be too bad, imo. Try Overleaf as a LaTeX "client", it does WYSIWYG very well. Excellent for collaboration too, if you need that. Lots of tutorials around you can google.
posted by MiraK at 6:15 PM on April 29, 2018 [8 favorites]

How do people handle this efficiently in their science/engineering papers?

Depends, but by far the best option is LaTeX. Your university probably already has template ready to go, off the shelf. There's some overhead in learning it, but stuff like numbering equations, citing them internally (and having it all work perfectly even if you reorder things) becomes trivially easy. Bibliography management is also a breeze.

Lots of people learn it for their dissertations, and the skills will pay off in the long run.

Seriously, it's so nice to be able to write science without wanting to strangle everyone due to frustration of shitty word processors. Word is in no way an adequate scientific manuscript tool, period. Do yourself a favor and use LaTeX, you won't regret it.
posted by SaltySalticid at 6:18 PM on April 29, 2018 [2 favorites]

I also came to ask if you must use word or if latex is an option?
posted by medusa at 6:24 PM on April 29, 2018

I have to use word for work, and I’ve tried really hard, but every method fails often enough to make the exercise futile. If you can’t just use LaTeX, like me, I’m sorry. What I do is label each equation latex/style, so equation {quadratic-whatever}, then refer to it later the same way. At the end, I use find/change to hand number. Restart numbering every section to not make yourself angry (lemma 5.1.2).

But i so miss typing \toc and having a perfectly formatted table of contents.
posted by Valancy Rachel at 6:51 PM on April 29, 2018 [2 favorites]

So this might be a clunky workaround, but the IDE I use in my day job allows you to write LaTeX syntax, click a button, and get a Word document out. The IDE is primarily for R and is called RStudio. I have never needed to number equations in it, but a little Googling makes me think it can be done.
posted by eirias at 6:58 PM on April 29, 2018

Just as a small warning: Word documents containing cross-references sometimes end up corrupted. When this happens, they can simply fail to save - silently, with no error message. Ever since this happened to me twice, I have numbered everything manually. I haven't found a better solution since I do not trust Word one fucking bit.

If you do use cross-references I suggest making sure you have frequent backups--I would go so far as to create a new copy each day you work on it.

It is the worst failure mode ever. You can be working on a document for hours without having any idea that it's not saving.
posted by Kutsuwamushi at 7:15 PM on April 29, 2018 [3 favorites]

If you really want to go hard core text processing, use Pandoc flavored markdown, then you can generate Word docx or LaTeX or or html, etc. It’s great stuff.

The world’s your oyster with pandoc*

*Or so I think, I haven’t moved to writing natively in it yet but it didn’t exist back when I was doing this and it might be an even better option than LaTeX at present because it sort of contains it and it will allow you to interface with goofs who demand Word files when you have to.
posted by SaltySalticid at 7:22 PM on April 29, 2018 [3 favorites]

(If you want to try to learn LaTeX—which is what I use too, but I’m a mathematician—send me a MeMail and I’ll send you the “getting started with LaTeX” stuff we’ve got floating around that we give to our students when we ask them to learn LaTeX in class.)
posted by leahwrenn at 7:43 PM on April 29, 2018

Seems like Ask has spoken, and we think you should learn LaTeX. I write for conference and journal submissions in a scientific field and I think you should learn LaTeX too. Managing long documents in Word is really challenging. Managing lists of equations or references is more challenging. Inserting figures without making your head explode is more challenging still. LaTeX + a decent class file and bibstyle makes all of these things nearly trivial, if you can learn some fairly minimal syntax.
posted by Alterscape at 7:46 PM on April 29, 2018

This does not exist. I'm sorry. People who are forced to use Word to manage equations (usually because our collaborators cannot or will not use LaTeX) simply must suffer. I know this will cost you hours of tedious and frustrating work. Again, I'm sorry.
posted by mr_roboto at 8:06 PM on April 29, 2018 [5 favorites]

Since it looks like the consensus is “learn LaTex,” I’m just dropping in to say there exist a variety of homegrown tools out there that exist to convert OCRed images to LaTex equations. Potentially Google may save you a lot of grief with equation formatting, at least.
posted by deludingmyself at 8:12 PM on April 29, 2018 [1 favorite]

So, i actually did this a few months ago (wrote my MEng thesis in Word) and... do not recommend it and am adding to the pileon that you should seriously consider LaTeX if it all possible. I went through a LOT of grief (including the save error Kutsuwamushi mentions above, which is really no fun) and my final archived, permanent digital copy has several "Error! Reference not found"s in the text because a bunch of references decided to spontaneously break right before I submitted. And this was for a masters and therefore shorter and much less involved than a doctoral diss... (if there is a specific reason you really, really need to use Word - the least awful solution I found was a custom caption template embedded into an invisible table for each equation. Memail me if you want more details, because I don't want to provide them here in an effort to encourage others who might stumble on this question to ditch Word. it's not worth it!!)
posted by btfreek at 8:42 PM on April 29, 2018 [3 favorites]

Best answer: I'm sorry that I can't help you answer the question you asked - I don't use Word for this type of thing.

You mentioned "two different computers and installs of MS Word 2016". If you're stuck using Word due to your collaborators, particularly if you love track changes (I do! I love Google Doc's version even more!), you might want to take a look at some nifty tools for collaborative, online LaTeX editing: ShareLatex or Overleaf. I personally prefer ShareLatex, but I think they're merging...
posted by Metasyntactic at 1:00 AM on April 30, 2018 [4 favorites]

Best answer: If your requirement for "Must work with Word 2016" is because your school requires the thesis to be submitted as a Word document, then you can always try this:

(1) Write it in LaTeX
(2) Take the completed file and process it through latex2rtf to generate a word-compatible file that will be super-ugly and have a few issues
(3) Take the rtf, fix the issues, make it pretty and compliant with university requirements, save as .docx or whatever.

If you keep trying to do this directly in Word, I will echo Kutsuwamushi's concern and urge you to have an extremely paranoid and comprehensive backup system. I myself lost a formal-theory paper when I opened the Word document and all the equations had turned into weird random dogshit that it wouldn't permit me to edit.
posted by GCU Sweet and Full of Grace at 4:28 AM on April 30, 2018 [1 favorite]

I will echo the advice that writing anything in Word that depends on Word's auto-numbering may end in tears.

I don't need to write scientific papers, but I do most of my writing in Markdown and process it through Pandoc to get whatever output format I need (you can do Markdown + math in Pandoc). If you need to submit a Word file specifically for some reason, I strongly recommend only using that as an output format, and not doing any real work in it apart from some cleanup formatting.
posted by adamrice at 8:10 AM on April 30, 2018 [1 favorite]

You know, LaTex fans can be super annoying their claims that you do ANYTHING with it, but, umm, they may actually be right about that.
posted by seasparrow at 12:53 PM on April 30, 2018 [2 favorites]

Response by poster: Thank you to everyone for your input. There are some very helpful suggestions here. These are the main takeaways I have from the responses:

1. This capability doesn't exist in MS Word and I am in for a nightmare heading down that path.
2. LaTex overwhelming is being encouraged as a solution.
3. Collaboration is important for me, so Overleaf or ShareTex are editors I should look into.
4. If necessary to convert to a Word format, there are ways to use rtf format as a bridge.

I will have to see if my advisor is on board with using collaborative tools other than MS Word markup features, but thanks again for the suggestions. I am going to dive into LaTex and see what I am missing.
posted by incolorinred at 6:39 AM on May 2, 2018 [1 favorite]