this is twice the same graph, except one is smoothed.
If you compile it, you will see.
documentclass[
]{scrreprt}
usepackage[backend=biber]{biblatex}
usepackage{pgfplots}
usepackage{tikz}
egin{document}
egin{figure*}[!ht]
centering
egin{tikzpicture}
egin{axis}[%
scatter/classes={%
1={mark=diamond*,draw=black},
0={mark=o,draw=black}},
yticklabels={},ytick style = {draw=none}, xticklabels={},xtick style = {draw=none},
xmin=0,
xmax=200,
ymin=0,
ymax=11,
width=12cm,
height=7cm]
addplot[smooth,black]
coordinates{
(0,10)
(10,1)
(200,0.1)
};
addplot[ dashed, black]
coordinates{
(0,10)
(10,1)
(200,0.1)
};
end{axis}
end{tikzpicture}
end{figure*}
end{document}
What I actually trying to achieve is that the second data point gives a smooth curve.
But using option smooth as you can see absolutely doesn't do the job here.
I tried with many more data points but nothing worked, this just gave me weird results. Also the approximation from the second to the last data point doesn't work as intended.
If you ask how it actually should look like, google Goodfellow generalization.
I want to draw this comparison between training and generalization error.
Every advice is very much appreciated!
question from:
https://stackoverflow.com/questions/65852459/problem-plotting-curve-with-smooth-option 与恶龙缠斗过久,自身亦成为恶龙;凝视深渊过久,深渊将回以凝视…