Gak: Modular Inverses!

[swestrup]

I hate the damn things. I know just enough number theory that I know when one would be useful, and then I take forever to calculate one. And Mathematica was of little help. I had to scour some web sites to determine that it uses a weird variant of its Solve function for linear congruence relations.

Here I was trying to get it to Solve[ Mod[17 a, 10000] == b, a ], which is the normal syntax (and which doesn't work) and it turns out I need to say Solve[ 17 a == b && Modulus == 10000, a], a fact which is not mentioned in the reference for EITHER Solve or Mod.

Comments

[sps.livejournal.com]

Don't you also need to say

... Answer->Correct, Please->Pretty, IMeanIt->Yes

???? Though I find it interesting that the first thing I tried to get it to do was a finite field problem and had this exact same problem. Mathematica is just soooo numerical in its outlook.

gaK, indeed!

[denizsarikaya.livejournal.com]

erm, weird. Matlab would have handled that in a fashion similar to the first incantation you wrote.

Of course, Matlab has its own quirks. Often, people just don't know there's a function which will do what they want (the day I learned about squeeze, I just about cried in pent-up frustration).