Where applicable, change pander etc. to kableextra.

In Adv Data Analysis, 'Closing Modeling Advice' in GAM chapter.

Note the viability of bam for very large data (esp. for RE).

Hi, everyone.
Thanks for this excelent tutorial for using gam.
I perfomed gam to my dataset testing the effect of phosphate addition in the growth of plants from three different locations.
I am wonder how to test for significant differences when comparing the three different locations.

Thanks in advance!

demo schoenberg package in technical section

Hello, Mr. Clark !

I just wanted to ask you if you'd be OK with me adapting your .css code for Bookdown for some of my projects. I referenced your work and your personal page in the .css file, but I can do it elsewhere too if you deem it appropriate.

In addition, I also added a couple of lines to "fix" some of what I considered small issues. I'll list them here if you're interested to incorporate them into your version.

• Fixed the width of horizontal lines
• Remove the font setting button
• Fix the width of sub elements in lists

In any case, I just wanted to say that your .css is really amazing to make Bookdown shine !

Update with content from stars workshop.

While plots are made with custom package (visibly), the visreg package is mentioned as an alternative. While it is viable, ggeffects is probably the better way to go and is ggplot by default.

Currently the reference section, while linked, is not separate from the appendix.

Also add Noam Ross's nifty interactive resource to it.
https://noamross.github.io/gams-in-r-course/

For the variance components table, the NA should be Residual, and no need to pipe for lmer_vcov. Maybe use gammit?

At the end, just for consistency leave out X_2

perhaps add a bit about time-varying data situations somewhere.