How do I copy and edit a Numbers file without changing the data?

Thanks Badunit for the close read. I will respond briefly, but there is good news. I think I “fixed” the problem, basically by duplicating the oddly behaving template file.

First about your point #6: it was in the process of preparing (with the read-only “git status” command) to commit my project (to a repository on GitHub) that I noticed the template file had a new modified date. But this could also be seen in a Finder window. This is before git changes anything.

I wanted to make a minimal failing case, so I took these steps:

1. make new folder on Desktop
2. option-drag in a duplicate of the template, rename it template.numbers
3. option-drag in a duplicate of a typical .cvs data file from my simulation, rename it new.cvs
4. ⌘ D duplicate template.numbers rename it new.numbers 
5. ⌘ O those two files (with Numbers) 
6. copy the data from new.csv then paste it into new.numbers
7. save new.numbers
8. quit Numbers

After step 4, the folder contained:

> ls -l
-rw-r--r--@ 1 cwr  staff    7157 Jun 16 16:15 new.csv
-rw-r--r--@ 1 cwr  staff  246999 Jun 20 10:08 new.numbers
-rw-r--r--@ 1 cwr  staff  246999 Jun 20 10:08 template.numbers

After step 8:

> ls -l
-rw-r--r--@ 1 cwr  staff    7157 Jun 16 16:15 new.csv
-rw-r--r--@ 1 cwr  staff  245977 Jun 21 13:42 new.numbers
-rw-r--r--@ 1 cwr  staff  246999 Jun 20 10:08 template.numbers

This is what I want!! The template.numbers file is unchanged (time 10:08) and new.numbers is updated (time 13:42) with the desired contents.

As unlikely as it sounds, perhaps this was some sort of transient corruption of that template file. Making a duplicate of the file might have fixed the issue.

The odd behavior was as if the files were aliased to each other. But I looked at the “bad template” with Finder's ⌘ I and it described it as a Numbers file not an alias.

So I think my problem is solved. And after helpful hints from SGIII, I better understand how to use the built-in templating in Numbers.

Thanks to you both.

Just a final note: I replaced my previous “bad template” with the new copy in my project's local git repository, then did a git commit to push that change to the server. I also replaced old with new in the Numbers template chooser. Everything seems to be working as expected now.

Thank you. I could have sworn I searched for a “Save as Template” in Numbers but obviously missed it.

What you suggest is an ideal “work around” for my problem. In fact I used the option to “Add to Template Chooser” which is more convenient than what I do now.

However I continue to see the odd “aliasing” behavior. When I create a new Numbers file from the newly saved template from my template chooser, it continues to modify the underlying file. I tested it a few minutes ago at 9:16am:

> ls -l plot_template.numbers
-rw-r--r--@ 1 cwr  staff  243426 Jun 20 09:16 plot_template.numbers

I think it is a workaround for the issue I actually raised. My original post was titled “copy Numbers file, edit copy, original is changed.” 

I appreciate you trying to help, and I did learn something about the Numbers template chooser. However that was not the issue I asked about. (Cartoon analogy: “Q: Every time I start my car, the windshield wipers start running. A: have you considered riding a bike?”)

To your questions: yes, it is significant and does affect the user. Files have had Date Modified fields since the 1960s because they are a key bit of file state. If a file named “2021 Financial report” has a last modified date of 2025, that is highly suspicious. When making a commit to a git repository, you are being asked to verify that each modified file was intentional, and to explain why. So every few days for the next 6 months, I will have to deal with this issue for each experiment I plot using this template.

What I hoped to learn is why this file is exhibiting this odd behavior. I know it is not typical. It may have nothing to do with Numbers. I just want to find a way to fix it.