Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > comp.soft-sys.math.mathematica > #10635 > unrolled thread

Higher precision in Error function Erf[] needed.

Started byCyril <cyril.stephanos@gmail.com>
First post2012-06-26 08:50 +0000
Last post2012-07-05 10:14 +0000
Articles 7 — 4 participants

Back to article view | Back to comp.soft-sys.math.mathematica


Contents

  Higher precision in Error function Erf[] needed. Cyril <cyril.stephanos@gmail.com> - 2012-06-26 08:50 +0000
    Re: Higher precision in Error function Erf[] needed. Ray Koopman <koopman@sfu.ca> - 2012-06-27 08:06 +0000
      Re: Higher precision in Error function Erf[] needed. Cyril <cyril.stephanos@gmail.com> - 2012-06-29 08:55 +0000
        Re: Higher precision in Error function Erf[] needed. Ray Koopman <koopman@sfu.ca> - 2012-06-30 09:17 +0000
          Re: Higher precision in Error function Erf[] needed. "Nasser M. Abbasi" <nma@12000.org> - 2012-07-02 09:28 +0000
      Re: Higher precision in Error function Erf[] needed. Ray Koopman <koopman@sfu.ca> - 2012-07-04 07:30 +0000
        Re: Higher precision in Error function Erf[] needed. "Alexey Popkov" <lehin.p@gmail.com> - 2012-07-05 10:14 +0000

#10635 — Higher precision in Error function Erf[] needed.

FromCyril <cyril.stephanos@gmail.com>
Date2012-06-26 08:50 +0000
SubjectHigher precision in Error function Erf[] needed.
Message-ID<jsbt4i$6pq$1@smc.vnet.net>
Hi everyone


I am using the error function as part of differential equations I am solving with NDSolve.
Unfortunately, for certain parameters, the error function Erf[] starts oscillating and yielding wrong values.

The oscillation does not show when I plot the error function itself, but when I plot this expression

Plot[Exp[8*(10 - Abs[10 - 20000*x])]*(1 - Erf[(8*(10 - Abs[10 - 20000*x]))^(1/2)]), {x, 0, 1*10^(-3)}, PlotRange -> Full]

it becomes visible.


When I do the integration for error function with NDSolve manually and plot the expression, the oscillation vanishes for high values of PrecisionGoal and WorkingPrecision, e.g.

Plot[ Exp[8*(10 - Abs[10 - 20000*x])]*(1 - NIntegrate[ 2/Sqrt[Pi] Exp[-t^2], {t, 0, (8*(10 - Abs[10 - 20000*x]))^(1/2)}, PrecisionGoal -> 40, WorkingPrecision -> 40]), {x, 0, 1*10^(-3)}, PlotRange -> Full]


The oscillation of Erf[] is not only visible in the plot, it also leads to wrong results in my calculations.

I tried to increase the precision of Erf[] by using N[Erf[],n] with very high values for n - as described in the Mathematica help - but it does not change anything.


Solving the Integral for the error function manually with NDSolve yields two problems. It is very slow - it takes about three minutes on my computer (I have to solve many equations) - whereas Erf[] is much faster. And it seems that I cannot use NDSolve as part of an equation I want to solve with NDSolve itself.

Is there a possibility to increase the precision of Erf[] and obtain more accurate results?


Thank you in advance for your help.


Best regards
Cyril

[toc] | [next] | [standalone]


#10659

FromRay Koopman <koopman@sfu.ca>
Date2012-06-27 08:06 +0000
Message-ID<jseeua$i3d$1@smc.vnet.net>
In reply to#10635
On Jun 26, 1:50 am, Cyril <cyril.stepha...@gmail.com> wrote:
> Hi everyone
>
> I am using the error function as part of differential equations I am solving with NDSolve.
> Unfortunately, for certain parameters, the error function Erf[] starts oscillating and yielding wrong values.
>
> The oscillation does not show when I plot the error function itself, but when I plot this expression
>
> Plot[Exp[8*(10 - Abs[10 - 20000*x])]*(1 - Erf[(8*(10 - Abs[10 - 20000*x]))^(1/2)]), {x, 0, 1*10^(-3)}, PlotRange -> Full]
>
> it becomes visible.
>
> When I do the integration for error function with NDSolve manually and plot the expression, the oscillation vanishes for high values of PrecisionGoal and WorkingPrecision, e.g.
>
> Plot[ Exp[8*(10 - Abs[10 - 20000*x])]*(1 - NIntegrate[ 2/Sqrt[Pi] Exp[-t^2], {t, 0, (8*(10 - Abs[10 - 20000*x]))^(1/2)}, PrecisionGoal -> 40, WorkingPrecision -> 40]), {x, 0, 1*10^(-3)}, PlotRange -> Full]
>
> The oscillation of Erf[] is not only visible in the plot, it also leads to wrong results in my calculations.
>
> I tried to increase the precision of Erf[] by using N[Erf[],n] with very high values for n - as described in the Mathematica help - but it does not change anything.
>
> Solving the Integral for the error function manually with NDSolve yields two problems. It is very slow - it takes about three minutes on my computer (I have to solve many equations) - whereas Erf[] is much faster. And it seems that I cannot use NDSolve as part of an equation I want to solve with NDSolve itself.
>
> Is there a possibility to increase the precision of Erf[] and obtain more accurate results?
>
> Thank you in advance for your help.
>
> Best regards
> Cyril

Change  (1 - Erf[...])  to  Erfc[...] .

[toc] | [prev] | [next] | [standalone]


#10691

FromCyril <cyril.stephanos@gmail.com>
Date2012-06-29 08:55 +0000
Message-ID<jsjqhg$d2j$1@smc.vnet.net>
In reply to#10659
On Wednesday, June 27, 2012 10:06:34 AM UTC+2, Ray Koopman wrote:
> On Jun 26, 1:50 am, Cyril <cyril.stepha...@gmail.com> wrote:
> > Hi everyone
> >
> > I am using the error function as part of differential equations I am solving with NDSolve.
> > Unfortunately, for certain parameters, the error function Erf[] starts oscillating and yielding wrong values.
> >
> > The oscillation does not show when I plot the error function itself, but when I plot this expression
> >
> > Plot[Exp[8*(10 - Abs[10 - 20000*x])]*(1 - Erf[(8*(10 - Abs[10 - 20000*x]))^(1/2)]), {x, 0, 1*10^(-3)}, PlotRange -> Full]
> >
> > it becomes visible.
> >
> > When I do the integration for error function with NDSolve manually and plot the expression, the oscillation vanishes for high values of PrecisionGoal and WorkingPrecision, e.g.
> >
> > Plot[ Exp[8*(10 - Abs[10 - 20000*x])]*(1 - NIntegrate[ 2/Sqrt[Pi] Exp[-t^2], {t, 0, (8*(10 - Abs[10 - 20000*x]))^(1/2)}, PrecisionGoal -> 40, WorkingPrecision -> 40]), {x, 0, 1*10^(-3)}, PlotRange -> Full]
> >
> > The oscillation of Erf[] is not only visible in the plot, it also leads to wrong results in my calculations.
> >
> > I tried to increase the precision of Erf[] by using N[Erf[],n] with very high values for n - as described in the Mathematica help - but it does not change anything.
> >
> > Solving the Integral for the error function manually with NDSolve yields two problems. It is very slow - it takes about three minutes on my computer (I have to solve many equations) - whereas Erf[] is much faster. And it seems that I cannot use NDSolve as part of an equation I want to solve with NDSolve itself.
> >
> > Is there a possibility to increase the precision of Erf[] and obtain more accurate results?
> >
> > Thank you in advance for your help.
> >
> > Best regards
> > Cyril
>
> Change  (1 - Erf[...])  to  Erfc[...] .


Ok, that was easy and works perfect. Thanks a lot!

Although it was not what I expected. Is there a possibility to alter the precision of Erf[]?

Best regards
Cyril

[toc] | [prev] | [next] | [standalone]


#10696

FromRay Koopman <koopman@sfu.ca>
Date2012-06-30 09:17 +0000
Message-ID<jsmg81$1jj$1@smc.vnet.net>
In reply to#10691
On Jun 29, 1:55 am, Cyril <cyril.stepha...@gmail.com> wrote:
> On Wednesday, June 27, 2012 10:06:34 AM UTC+2, Ray Koopman wrote:
>> On Jun 26, 1:50 am, Cyril <cyril.stepha...@gmail.com> wrote:
>>> Hi everyone
>>>
>>> I am using the error function as part of differential equations I
>>> am solving with NDSolve. Unfortunately, for certain parameters, the
>>> error function Erf[] starts oscillating and yielding wrong values.
>>>
>>> The oscillation does not show when I plot the error function itself,
>>> but when I plot this expression
>>>
>>> Plot[Exp[8*(10 - Abs[10 - 20000*x])]*(1 - Erf[(8*(10 - Abs[10 -
>>> 20000*x]))^(1/2)]), {x, 0, 1*10^(-3)}, PlotRange -> Full]
>>>
>>> it becomes visible.
>>>
>>> When I do the integration for error function with NDSolve manually
>>> and plot the expression, the oscillation vanishes for high values
>>> of PrecisionGoal and WorkingPrecision, e.g.
>>>
>>> Plot[ Exp[8*(10 - Abs[10 - 20000*x])]*(1 - NIntegrate[ 2/Sqrt[Pi]
>>> Exp[-t^2], {t, 0, (8*(10 - Abs[10 - 20000*x]))^(1/2)},
>>> PrecisionGoal -> 40, WorkingPrecision -> 40]), {x, 0, 1*10^(-3)},
>>> PlotRange -> Full]
>>>
>>> The oscillation of Erf[] is not only visible in the plot,
>>> it also leads to wrong results in my calculations.
>>>
>>> I tried to increase the precision of Erf[] by using N[Erf[],n]
>>> with very high values for n - as described in the Mathematica help
>>> - but it does not change anything.
>>>
>>> Solving the Integral for the error function manually with NDSolve
>>> yields two problems. It is very slow - it takes about three minutes
>>> on my computer (I have to solve many equations) - whereas Erf[] is
>>> much faster. And it seems that I cannot use NDSolve as part of an
>>> equation I want to solve with NDSolve itself.
>>>
>>> Is there a possibility to increase the precision of Erf[]
>>> and obtain more accurate results?
>>>
>>> Thank you in advance for your help.
>>>
>>> Best regards
>>> Cyril
>>
>> Change  (1 - Erf[...])  to  Erfc[...] .
>
> Ok, that was easy and works perfect. Thanks a lot!
>
> Although it was not what I expected.
> Is there a possibility to alter the precision of Erf[]?
>
> Best regards
> Cyril

Just alter the precision of its input. First define

  f[x_,n_] := 8*(10 - Abs[10 - 20000*SetPrecision[x,n]])

Then

  Plot[Exp@#*(1-Erf@Sqrt@#)&@f[x,20], {x,0,10^-3}, Compiled->False]

works fine.

[toc] | [prev] | [next] | [standalone]


#10739

From"Nasser M. Abbasi" <nma@12000.org>
Date2012-07-02 09:28 +0000
Message-ID<jsrpkh$d3i$1@smc.vnet.net>
In reply to#10696
On 6/30/2012 4:17 AM, Ray Koopman wrote:

>
> Just alter the precision of its input. First define
>
>    f[x_,n_] := 8*(10 - Abs[10 - 20000*SetPrecision[x,n]])
>
> Then
>
>    Plot[Exp@#*(1-Erf@Sqrt@#)&@f[x,20], {x,0,10^-3}, Compiled->False]
>
> works fine.
>

fyi;

using Mathematica 8, the 'Compiled->False' is shown in RED
letters, meaning is not valid option for Plot.

on a side note, I myself, and this is just a styling thing,
when using pure function, find it easier to use @ for
only the RHS of the pure function. Hence the above command
would become

  Plot[ Exp[#] * (1-Erf[Sqrt[#]])& @f[x,20], {x,0,10^-3}]

vs.

  Plot[Exp@# * (1-Erf@Sqrt@#)& @f[x,20], {x,0,10^-3}]

Again, just different styling, all is valid, but for me, I
find the first case above a little easier to read, since @ is
used only for the pure function argument.

regards,
--Nasser

[toc] | [prev] | [next] | [standalone]


#10773

FromRay Koopman <koopman@sfu.ca>
Date2012-07-04 07:30 +0000
Message-ID<jt0rdo$1ah$1@smc.vnet.net>
In reply to#10659
On Jul 2, 2:28 am, "Nasser M. Abbasi" <n...@12000.org> wrote:
> On 6/30/2012 4:17 AM, Ray Koopman wrote:
>
>> Just alter the precision of its input. First define
>>
>>    f[x_,n_] := 8*(10 - Abs[10 - 20000*SetPrecision[x,n]])
>>
>> Then
>>
>>    Plot[Exp@#*(1-Erf@Sqrt@#)&@f[x,20], {x,0,10^-3}, Compiled->False]
>>
>> works fine.
>
> fyi;
>
> using Mathematica 8, the 'Compiled->False' is shown in RED
> letters, meaning is not valid option for Plot.

So the advice at
<http://reference.wolfram.com/mathematica/ref/Compiled.html>
to use Compiled->False when plotting is "no longer operative"?

[toc] | [prev] | [next] | [standalone]


#10798

From"Alexey Popkov" <lehin.p@gmail.com>
Date2012-07-05 10:14 +0000
Message-ID<jt3pes$12r$1@smc.vnet.net>
In reply to#10773
I just tried  both variants:

Plot[If[Exp[-x] == 0., 0, -1] + x/900, {x, 0, 1000}, Compiled -> False]

Plot[If[Exp[-x] == 0., 0, -1] + x/900, {x, 0, 1000},
Method -> {Compiled -> False}]

In the first case Compiled is shown in red but it works as it should 
according to the documentation page you referenced. But the second method 
suggested on that documentation page does not work.

P.S. Using Trace, I discovered potentially useful error message of Plot:

On[Plot::accbend]
Plot[If[Exp[-x] == 0., 0, -1] + x/900, {x, 0, 1000}]

(*=> Plot::accbend: Target bending angle of 4.999999999999992` not 
satisfied; current maximal bending angle is 72.68343019044056`. >>*)


"Ray Koopman" <koopman@sfu.ca> wrote: news:jt0rdo$1ah$1@smc.vnet.net...
> On Jul 2, 2:28 am, "Nasser M. Abbasi" <n...@12000.org> wrote:
>> On 6/30/2012 4:17 AM, Ray Koopman wrote:
>>
>>> Just alter the precision of its input. First define
>>>
>>>    f[x_,n_] := 8*(10 - Abs[10 - 20000*SetPrecision[x,n]])
>>>
>>> Then
>>>
>>>    Plot[Exp@#*(1-Erf@Sqrt@#)&@f[x,20], {x,0,10^-3}, Compiled->False]
>>>
>>> works fine.
>>
>> fyi;
>>
>> using Mathematica 8, the 'Compiled->False' is shown in RED
>> letters, meaning is not valid option for Plot.
>
> So the advice at
> <http://reference.wolfram.com/mathematica/ref/Compiled.html>
> to use Compiled->False when plotting is "no longer operative"?
> 


[toc] | [prev] | [standalone]


Back to top | Article view | comp.soft-sys.math.mathematica


csiph-web