why the curve for smoothing term different?

According to what i had read,that An edf of 2 is equivalent to a quadratic curve, but my plot does not look like quadratic curve. i want to know the reason. This the link for that i refer to https://noamross.github.io/gams-in-r-course/chapter2

I wanted to add another question, what make plot curve in fb.ratio different from SOM even when both had edf=2?

It is because of the distribution for observation data point in fb.ratio more disperse while SOM observation more collected in group?

• Please don't edit questions to add new questions. It's better to have a single question per thread, and substantial edits to the question will invalidate existing answers. Instead. please ask a new self-contained question. (Do feel free to link here from the new question, and from this one to the new one, to provide context.) Jun 16, 2020 at 6:10

It's all a question of scaling the vertical axis. Here are two plots of $$y=x^2$$. In the left one, the $$y$$ axis is scaled so the curve fills the vertical space, in the right one the $$y$$ axis is scaled (and the $$x$$ axis is truncated) to give something similar to your data - but both plots show the same parabola:
xx <- seq(-1,1,.0001)