Mathematica 9 is now available
Student Support Forum
Student Support Forum > General > > "Strange NdSolve solution"

Post Reply:
Email Address:

Please send email when my message is replied to.

Url (optional):
Message: view original message?
Attachment (optional):
Please answer this:2+5 =

Original Message (ID '429360') By Jigoro:
In Response To 'Re: Strange NdSolve solution' --------- In fact, if we write a code as follows: y[t_]=Exp[2*t]+2*Exp[-t] s=NDSolve[{z"''"[t]-z'[t]-2*z[t]==0,z[0]==3,z'[0]==0},z,{t,-100,100}] Then we can see, for example, that y[10]- z[10]/.s[[1]] is about 124, and not close to zero. Percentage wise, the discrepancy is not big, but 124 is not a “small” number either. I wonder how to make differences as close to zero as possible. Thanks.