Error with floats in implementation of Horner's method. - Beginners - Haskell.org

newer
strip --strip-unneeded on package...

Error with floats in implementation of Horner's method.

older
Cabal build dependency using a...

Andrew Fleckenstein

13 Dec 2013 13 Dec '13

9:03 p.m.

Hi all, I have implemented Horner's method for evaluating a polynomial as follows: -- the nth horner element, given a list of coefficients and a value at which it needs to be evaluated horner_element l x n | n == (length l)-1 = last l | otherwise = (l !! n) + (horner_element l x (n+1))*x -- compute a polynomial given as a list of coefficients at the value x using horner's method. horner l x = horner_element l x 0

Attachments:

attachment.html (text/html — 501 bytes)

Reply

Sign in to reply online Use email software

Show replies by date

Andrew Fleckenstein

13 Dec 13 Dec

9:07 p.m.

New subject: Error with floats in implementation of Horner's method.

And I have sent the email before it was finished, oops. Evaluating at integer values works fine, but trying to do *Main> let a = [-1,2,-6,2] *Main> horner a 1.23 gives this error <interactive>:4:10: No instance for (Fractional Integer) arising from the literal `1.23' Possible fix: add an instance declaration for (Fractional Integer) In the second argument of `horner', namely `1.23' In the expression: horner a 1.23 In an equation for `it': it = horner a 1.23 I know it has something to do with types, but whenever I try to add a type signature to the functions it just messes everything up even more. Any help would be appreciated Thanks, Andrew On Fri, Dec 13, 2013 at 4:03 PM, Andrew Fleckenstein < andrew.fleckenstein@gmail.com> wrote:

Hi all,

I have implemented Horner's method for evaluating a polynomial as follows:

-- the nth horner element, given a list of coefficients and a value at which it needs to be evaluated

horner_element l x n | n == (length l)-1 = last l | otherwise = (l !! n) + (horner_element l x (n+1))*x

-- compute a polynomial given as a list of coefficients at the value x using horner's method.

horner l x = horner_element l x 0

Reply

Sign in to reply online Use email software

Kim-Ee Yeoh

9:20 p.m.

New subject: Error with floats in implementation of Horner's method.

On Sat, Dec 14, 2013 at 4:07 AM, Andrew Fleckenstein < andrew.fleckenstein@gmail.com> wrote:

I know it has something to do with types,

Right! :)

but whenever I try to add a type signature to the functions it just messes everything up even more.

This works: let a :: [Double]; a = [-1,2,-6,2] This works too: let a :: Num a => [a]; a = [-1,2,-6,2] What do you think are the pros and cons of either? -- Kim-Ee

Reply

Sign in to reply online Use email software

Andrew Fleckenstein

9:51 p.m.

New subject: Error with floats in implementation of Horner's method.

I know Num is a class, and Double is an instance of the Num class. I guess using Num would make it more generic, but I don't see why either one would be better in this particular function. On Fri, Dec 13, 2013 at 4:20 PM, Kim-Ee Yeoh wrote:

On Sat, Dec 14, 2013 at 4:07 AM, Andrew Fleckenstein < andrew.fleckenstein@gmail.com> wrote:

...
I know it has something to do with types,

Right! :)

...
but whenever I try to add a type signature to the functions it just messes everything up even more.

This works: let a :: [Double]; a = [-1,2,-6,2]

This works too: let a :: Num a => [a]; a = [-1,2,-6,2]

What do you think are the pros and cons of either?

-- Kim-Ee

_______________________________________________ Beginners mailing list Beginners@haskell.org http://www.haskell.org/mailman/listinfo/beginners

Reply

Sign in to reply online Use email software

Brent Yorgey

14 Dec 14 Dec

2:21 a.m.

New subject: Error with floats in implementation of Horner's method.

Right. The problem is that when you write let a = [1,2,3,4] at the GHCi prompt, the Evil Monomorphism Restriction forces it to have a monomorphic type, and GHCi picks [Integer]. I recommend that you turn off the Monomorphism Restriction, which you can do in one of the following ways: 1) Type :set -XNoMonomorphismRestriction at the GHCi prompt 2) Add :set -XNoMonomorphismRestriction to your ~/.ghci file. If you do the latter then you only have to do it once, and from then on the MR will always be turned off in GHCi. With the MR turned off, let cs = [1,2,3,4] will result in cs :: Num a => [a] which will work the way you expect. -Brent On Fri, Dec 13, 2013 at 04:51:25PM -0500, Andrew Fleckenstein wrote:

I know Num is a class, and Double is an instance of the Num class. I guess using Num would make it more generic, but I don't see why either one would be better in this particular function.

On Fri, Dec 13, 2013 at 4:20 PM, Kim-Ee Yeoh wrote:

...
On Sat, Dec 14, 2013 at 4:07 AM, Andrew Fleckenstein < andrew.fleckenstein@gmail.com> wrote:

...
I know it has something to do with types,

Right! :)

...
but whenever I try to add a type signature to the functions it just messes everything up even more.

This works: let a :: [Double]; a = [-1,2,-6,2]

This works too: let a :: Num a => [a]; a = [-1,2,-6,2]

What do you think are the pros and cons of either?

-- Kim-Ee

_______________________________________________ Beginners mailing list Beginners@haskell.org http://www.haskell.org/mailman/listinfo/beginners

_______________________________________________ Beginners mailing list Beginners@haskell.org http://www.haskell.org/mailman/listinfo/beginners

Reply

Sign in to reply online Use email software

4220

Age (days ago)

4221

Last active (days ago)

Download

4 comments

3 participants

tags

participants (3)

Andrew Fleckenstein
Brent Yorgey
Kim-Ee Yeoh