Do a final build without the cache

[gavinsimpson]

I was hitting issues running the code for the basis function plot, which may not be showing up locally as the figure is cached.

Suggest hat once the final version is ready for resubmission that we delete the cache and re-run the entire build process to be sure that all code blocks run fine.

[eric-pedersen]

I think I've figured this out in the gratia branch, as it wasn't stated in the build file that the basis function code was required (now changed). We should each try to build it from scratch though, before final submission, just to make sure it runs on different machines.

[gavinsimpson]

@eric-pedersen Thanks for merging everything in the various PRs. I'm now doing a full build locally having deleted all the cached stuff. I'll let you know in a wee while how I get on with it.

[eric-pedersen]

Great! I just made a final push to the repository, tweaking the makefile to create a pdf with text differences in it, but it shouldn't affect your build (there's a switch in it that has to be changed to true to compile the track-changes version).

Let me know how it runs.

[eric-pedersen]

Did the run work? Did we get a clean build of health? 😄

[dill]

Would it be useful for me to do this on my Mac too?

On 09/01/2019 05:22, Eric Pedersen wrote:

Did the run work? Did we get a clean build of health? 😄

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/noamross/mixed-effect-gams/issues/44#issuecomment-452575894, or mute the thread https://github.com/notifications/unsubscribe-auth/AAAHoZ-akm8EnqPlEH_HGzkKzQ2BnbYuks5vBXybgaJpZM4ZlG4M.

[gavinsimpson]

Did the run work? Did we get a clean build of health?

Yes, but I'm seeing some rendering issues esp section headers for parts 1 and 2. iI also saw 30+ warnings reported that I'd like to check the cuase of.

[eric-pedersen]

Would it be useful for me to do this on my Mac too?

I think it would be good. Better to make sure it compiles under multiple environments.

[eric-pedersen]

Did the run work? Did we get a clean build of health?

Yes, but I'm seeing some rendering issues esp section headers for parts 1 and 2. iI also saw 30+ warnings reported that I'd like to check the cause of.

Hhhmm, strange. Could you post a screenshot of the mis-rendered headers?

[gavinsimpson]

Hhhmm, strange. Could you post a screenshot of the mis-rendered headers?

I think I know what the issue is (need a space between # and the start of the header) but I'm just checking. If so this is just differences in terms of what our respective pandoc installs/versions allow.

[gavinsimpson]

@eric-pedersen Yep; that's fixed that rendering issue. I'll push those changes now.

[dill]

On the off-chance any of you have encountered this error before...

I am able to get from Rmd -> md -> tex but then get:

Output created: full_document.tex
sed -i '/xcolor/d' paper_sections/full_document.tex
sed: 1: "paper_sections/full_doc ...": extra characters at the end of p 
command
make: *** [main] Error 1

On 09/01/2019 15:52, Eric Pedersen wrote:

Would it be useful for me to do this on my Mac too?

I think it would be good. Better to make sure it compiles under multiple environments.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/noamross/mixed-effect-gams/issues/44#issuecomment-452746197, or mute the thread https://github.com/notifications/unsubscribe-auth/AAAHoX-I4UpDGwdyQWd9tbFtrpKZB_GCks5vBhBbgaJpZM4ZlG4M.

[gavinsimpson]

@dill Does it work if the sed command in the Makefile is changed to:

sed -i '' '/xcolor/d' paper_sections/full_document.tex

I.e. a separate set of single quotes before the actual command.

This is from: https://stackoverflow.com/questions/16745988/sed-command-with-i-option-in-place-editing-works-fine-on-ubuntu-but-not-mac

[dill]

Should have googled that myself! Thanks for the pointer.

Looks like the "best" solution would be to do what is described here:

https://stackoverflow.com/a/20951570

(that leaves our other BSD unices, but my guess is they have minimal following of our paper compared to linux/mac/windows :))

Alternatively we can ignore this or put something in the README? (Or maybe a Makefile.mac for folks to use?

On 10/01/2019 15:52, Gavin Simpson wrote:

@dill https://github.com/dill Does it work if the |sed| command in the |Makefile| is changed to:

|sed -i '' '/xcolor/d' paper_sections/full_document.tex |

I.e. a separate set of single quotes before the actual command.

This is from: https://stackoverflow.com/questions/16745988/sed-command-with-i-option-in-place-editing-works-fine-on-ubuntu-but-not-mac

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/noamross/mixed-effect-gams/issues/44#issuecomment-453145591, or mute the thread https://github.com/notifications/unsubscribe-auth/AAAHof0iPNzrkzV1ezOsEkNVAJ3H_24tks5vB2HRgaJpZM4ZlG4M.

[dill]

Also, I should say, the paper looks good to me!

On 10/01/2019 15:52, Gavin Simpson wrote:

@dill https://github.com/dill Does it work if the |sed| command in the |Makefile| is changed to:

|sed -i '' '/xcolor/d' paper_sections/full_document.tex |

I.e. a separate set of single quotes before the actual command.

This is from: https://stackoverflow.com/questions/16745988/sed-command-with-i-option-in-place-editing-works-fine-on-ubuntu-but-not-mac

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/noamross/mixed-effect-gams/issues/44#issuecomment-453145591, or mute the thread https://github.com/notifications/unsubscribe-auth/AAAHof0iPNzrkzV1ezOsEkNVAJ3H_24tks5vB2HRgaJpZM4ZlG4M.

[eric-pedersen]

Ah, the joys of of command line bullshittery... http://www.pgbovine.net/command-line-bullshittery.htm

Thanks for digging into this, and trying to find a fix. My preferred solution would be to have one makefile that works cross-platform. @dill Do you want to try the fix listed at the Stack Overflow site, and see if it works, then push a branch so Gavin and I can test if it compiles as well?

[dill]

I tried a potentially easier solution, if someone else can try it on their machine that would be ace.

[eric-pedersen]

I just tried a clean build on my work computer (Windows 2010), and it worked fine for me! I'll try again tonight on my personal computer (Windows 10), but I think this solution seems to be working.

[dill]

Ace! Glad to hear it!

On 10/01/2019 21:17, Eric Pedersen wrote:

I just tried a clean build on my work computer (Windows 2010), and it worked fine for me! I'll try again tonight on my personal computer (Windows 10), but I think this solution seems to be working.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/noamross/mixed-effect-gams/issues/44#issuecomment-453257191, or mute the thread https://github.com/notifications/unsubscribe-auth/AAAHoRkNaclVERuzguq-hfs58Fe1GIoPks5vB635gaJpZM4ZlG4M.

[gavinsimpson]

That worked for me on Linux; I didn't do a clean build as we're only testing what happens with the .tex part, but it all looks fine to me too.

[eric-pedersen]

And it works fine on Windows 10 as well! I'd say feel free to go ahead and pull this into master.

[dill]

Done!