MuPAD w. MATLAB 2015a: How to (properly) plot x*exp(-x/1.e-8) ?

I tried the following plot command in MuPAD running in MATLAB 2015a:

plot(x*exp(-x/1.e-8), x=0..5e-8)

and also tried:

plot(x*exp(-x/1.e-8), ViewingBox=[0..5e-8, 0..5e-9])

The result in both cases is a vertical axis from -1 to 1 and a graph that traces y=0. These commands properly displayed the graph in the MATLAB 2013a version of MuPAD (the maximum value of the function is approx. 3.7e-9 occurring at x approx. 1.e-8. Another thing I discovered was that if I use 1.e-7 in the function instead of 1.e-8, the 2015a version works fine. How can it properly plot it using 1.e-8 (or smaller) using the 2015a version of MuPAD?

[Walter Roberson]

I wonder if it would make a difference if you increased Digits ?

[Michael Pangia]

Thanks for the good suggestion, Walter!

I just tried it, and it doesn't correct the plotting problem.

Specifically, I tried:

DIGITS:=50

I also tested the numerical accuracy of a relevant calculation:

x*exp(-x/1.e-8) |x=1.e-8

and got the correct answer (to 50 significant figures):

0.0000000036787944117144232159552377016146086744581113103177

If it's a computational-accuracy error, it would seem to be happening within the plot command. Still puzzled ...

[Walter Roberson]

Out of curiosity, how about if you switch it to an explicit multiply inside of divide by a fraction?

plot(x*exp(-x*1.e8), x=0..5e-8)

[Michael Pangia]

Just tried your suggestion:

plot(x*exp(-x*1.e8), x=0..5e-8)

Same basic problem. The plot traces y=0 using a vertical range of -1 to 1, but if I use 1.e7, as in the following, that function gets properly displayed using a vertical range of 0 to 4e-8

plot(x*exp(-x*1.e7),x=0..5e-7)