Rocksolid Light

Welcome to novaBBS (click a section below)

mail  files  register  newsreader  groups  login

Message-ID:  

fortune: not found


devel / comp.theory / Anyone got a nice S-shaped function?

SubjectAuthor
* Anyone got a nice S-shaped function?Malcolm McLean
+* Anyone got a nice S-shaped function?John Forkosh
|`- Anyone got a nice S-shaped function?Malcolm McLean
+* Anyone got a nice S-shaped function?B.H.
|`* Anyone got a nice S-shaped function?B.H.
| `- Anyone got a nice S-shaped function?B.H.
`- Anyone got a nice S-shaped function?Keith Thompson

1
Anyone got a nice S-shaped function?

<f9dc1b80-9215-44ed-802f-98922df6d265n@googlegroups.com>

 copy mid

https://www.novabbs.com/devel/article-flat.php?id=36718&group=comp.theory#36718

 copy link   Newsgroups: comp.theory
X-Received: by 2002:a05:6214:2aac:b0:474:8b64:8f56 with SMTP id js12-20020a0562142aac00b004748b648f56mr5023759qvb.0.1659697142136;
Fri, 05 Aug 2022 03:59:02 -0700 (PDT)
X-Received: by 2002:a81:8413:0:b0:322:d661:a785 with SMTP id
u19-20020a818413000000b00322d661a785mr5237055ywf.16.1659697141907; Fri, 05
Aug 2022 03:59:01 -0700 (PDT)
Path: i2pn2.org!i2pn.org!usenet.blueworldhosting.com!feed1.usenet.blueworldhosting.com!peer01.iad!feed-me.highwinds-media.com!news.highwinds-media.com!news-out.google.com!nntp.google.com!postnews.google.com!google-groups.googlegroups.com!not-for-mail
Newsgroups: comp.theory
Date: Fri, 5 Aug 2022 03:59:01 -0700 (PDT)
Injection-Info: google-groups.googlegroups.com; posting-host=2a00:23a8:400a:5601:fc32:afe8:789f:1a87;
posting-account=Dz2zqgkAAADlK5MFu78bw3ab-BRFV4Qn
NNTP-Posting-Host: 2a00:23a8:400a:5601:fc32:afe8:789f:1a87
User-Agent: G2/1.0
MIME-Version: 1.0
Message-ID: <f9dc1b80-9215-44ed-802f-98922df6d265n@googlegroups.com>
Subject: Anyone got a nice S-shaped function?
From: malcolm....@gmail.com (Malcolm McLean)
Injection-Date: Fri, 05 Aug 2022 10:59:02 +0000
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
X-Received-Bytes: 2053
 by: Malcolm McLean - Fri, 5 Aug 2022 10:59 UTC

I'm looking for a nice S-shaped function which is zero at x= 00, 1 at x = 1, and is montonically increasing.
I want it to be possible to make it very close to the sum of absolute derivatives of the cosine curve (this is the curve you obtain by placing marks at even spaces on the edge of a semicircle, then take the x-coordinates), and I want it to respond intuitively to two markers I'll give to the user. The mid point of the two markers should be the inflection point of the curve, the distance from the marker to the x=0 or x =1 point represents the steepness of the curve.

(What I want it for is to allow the user to specify curved surfaces without having a full 3D drawing system. Most of those surfaces will be semicircular because he'll be drawing a can or a bottle or something similar. Hence the cosine-based curve will be the default. But I want him to have a bit of control
to tweak the surface).

Re: Anyone got a nice S-shaped function?

<tckpp2$7oe$1@reader2.panix.com>

 copy mid

https://www.novabbs.com/devel/article-flat.php?id=36777&group=comp.theory#36777

 copy link   Newsgroups: comp.theory
Path: i2pn2.org!i2pn.org!weretis.net!feeder6.news.weretis.net!panix!.POSTED.panix3.panix.com!not-for-mail
From: fork...@panix.com (John Forkosh)
Newsgroups: comp.theory
Subject: Re: Anyone got a nice S-shaped function?
Date: Sat, 6 Aug 2022 04:13:22 -0000 (UTC)
Organization: PANIX Public Access Internet and UNIX, NYC
Message-ID: <tckpp2$7oe$1@reader2.panix.com>
References: <f9dc1b80-9215-44ed-802f-98922df6d265n@googlegroups.com>
Injection-Date: Sat, 6 Aug 2022 04:13:22 -0000 (UTC)
Injection-Info: reader2.panix.com; posting-host="panix3.panix.com:166.84.1.3";
logging-data="7950"; mail-complaints-to="abuse@panix.com"
User-Agent: tin/2.6.0-20210823 ("Coleburn") (NetBSD/9.2 (amd64))
 by: John Forkosh - Sat, 6 Aug 2022 04:13 UTC

Malcolm McLean <malcolm.arthur.mclean@gmail.com> wrote:
> I'm looking for a nice S-shaped function which is zero at x= 00,
> 1 at x = 1, and is montonically increasing.
> I want it to be possible to make it very close to the sum
> of absolute derivatives of the cosine curve (this is the curve
> you obtain by placing marks at even spaces on the edge of a
> semicircle, then take the x-coordinates), and I want it to
> respond intuitively to two markers I'll give to the user.
> The mid point of the two markers should be the inflection point
> of the curve, the distance from the marker to the x=0 or x =1 point
> represents the steepness of the curve.
>
> (What I want it for is to allow the user to specify curved surfaces
> without having a full 3D drawing system. Most of those surfaces
> will be semicircular because he'll be drawing a can or a bottle
> or something similar. Hence the cosine-based curve will be the default.
> But I want him to have a bit of control to tweak the surface).

I'm not following "S-shaped" and "monotonically increasing";
sounds contradictory. In any case, have you considered
Bezier curves,
https://en.wikipedia.org/wiki/B%C3%A9zier_curve
The first and last control points become the corresponding
endpoints of the generated bezier curve, so specifying them
as (0,0),(1,1) satisfies your boundary conditions.
And you can use whatever input the user's providing to
determine as many intermediate control points as you want,
and/or let the user explicitly specify them.
Here's a short routine I use for bezier curves, where
"point" is the obvious struct, and "binky(n,k)" is the
binomial coefficient n!/k!(n-k)! ...

/*===========================================================================
* Function: bezier ( n, cp, t )
* Purpose: returns x,y-coords of bezier curve
* with n control points cp, at 0.<=t<=1.
* --------------------------------------------------------------------------
* Arguments: n (I) int n>=2, number of control points
* cp (I) point* vector of control points, with
* cp[0] the initial, and cp[n-1] the final
* points of the curve, as usual.
* t (I) double 0.<=t<=1.
* Returns: (point) x,y-coords of bezier curve at t,
* or -1.,-1. for argument error, or any error.
* --------------------------------------------------------------------------
* Notes: o
*======================================================================== */
/* --- entry point --- */
point bezier ( int n, point *cp, double t ) {
/* --- allocations and declarations --- */
point bezt = { -1., -1. }; /* returned point, init for error */
int k=0, binky(); /* binomial coefficient */
/* --- check args --- */
if ( n<2 || cp==NULL /* invalid control points */
|| t<0. || t>1. ) goto end_of_job; /* invalid parameter */
/* --- explicit sum over control points --- */
n--; /* sum from 0...n rather than 0...n-1 */
bezt.x = bezt.y = 0.0; /* init sum */
for ( k=0; k<=n; k++ ) {
double coef=((double)binky(n,k))*pow(1.-t,(double)(n-k))*pow(t,(double)k);
bezt.x += coef*cp[k].x;
bezt.y += coef*cp[k].y;
} /* --- end-of-for(k) --- */
end_of_job:
return ( bezt ); /* back to caller */
} /* --- end-of-function bezier() --- */

--
John Forkosh ( mailto: j@f.com where j=john and f=forkosh )

Re: Anyone got a nice S-shaped function?

<1dc2e79d-f21c-45c0-b110-b108a1ab4203n@googlegroups.com>

 copy mid

https://www.novabbs.com/devel/article-flat.php?id=36787&group=comp.theory#36787

 copy link   Newsgroups: comp.theory
X-Received: by 2002:a05:622a:174c:b0:31f:3926:c809 with SMTP id l12-20020a05622a174c00b0031f3926c809mr9116727qtk.579.1659786748108;
Sat, 06 Aug 2022 04:52:28 -0700 (PDT)
X-Received: by 2002:a81:53d6:0:b0:31c:c750:14f9 with SMTP id
h205-20020a8153d6000000b0031cc75014f9mr10152143ywb.248.1659786747743; Sat, 06
Aug 2022 04:52:27 -0700 (PDT)
Path: i2pn2.org!i2pn.org!usenet.blueworldhosting.com!feed1.usenet.blueworldhosting.com!peer03.iad!feed-me.highwinds-media.com!news.highwinds-media.com!news-out.google.com!nntp.google.com!postnews.google.com!google-groups.googlegroups.com!not-for-mail
Newsgroups: comp.theory
Date: Sat, 6 Aug 2022 04:52:27 -0700 (PDT)
In-Reply-To: <tckpp2$7oe$1@reader2.panix.com>
Injection-Info: google-groups.googlegroups.com; posting-host=2a00:23a8:400a:5601:bc8c:d2bc:4bd0:86d0;
posting-account=Dz2zqgkAAADlK5MFu78bw3ab-BRFV4Qn
NNTP-Posting-Host: 2a00:23a8:400a:5601:bc8c:d2bc:4bd0:86d0
References: <f9dc1b80-9215-44ed-802f-98922df6d265n@googlegroups.com> <tckpp2$7oe$1@reader2.panix.com>
User-Agent: G2/1.0
MIME-Version: 1.0
Message-ID: <1dc2e79d-f21c-45c0-b110-b108a1ab4203n@googlegroups.com>
Subject: Re: Anyone got a nice S-shaped function?
From: malcolm....@gmail.com (Malcolm McLean)
Injection-Date: Sat, 06 Aug 2022 11:52:28 +0000
Content-Type: text/plain; charset="UTF-8"
X-Received-Bytes: 5447
 by: Malcolm McLean - Sat, 6 Aug 2022 11:52 UTC

On Saturday, 6 August 2022 at 05:13:25 UTC+1, John Forkosh wrote:
> Malcolm McLean <malcolm.ar...@gmail.com> wrote:
> > I'm looking for a nice S-shaped function which is zero at x= 00,
> > 1 at x = 1, and is montonically increasing.
> > I want it to be possible to make it very close to the sum
> > of absolute derivatives of the cosine curve (this is the curve
> > you obtain by placing marks at even spaces on the edge of a
> > semicircle, then take the x-coordinates), and I want it to
> > respond intuitively to two markers I'll give to the user.
> > The mid point of the two markers should be the inflection point
> > of the curve, the distance from the marker to the x=0 or x =1 point
> > represents the steepness of the curve.
> >
> > (What I want it for is to allow the user to specify curved surfaces
> > without having a full 3D drawing system. Most of those surfaces
> > will be semicircular because he'll be drawing a can or a bottle
> > or something similar. Hence the cosine-based curve will be the default.
> > But I want him to have a bit of control to tweak the surface).
> I'm not following "S-shaped" and "monotonically increasing";
> sounds contradictory. In any case, have you considered
> Bezier curves,
> https://en.wikipedia.org/wiki/B%C3%A9zier_curve
> The first and last control points become the corresponding
> endpoints of the generated bezier curve, so specifying them
> as (0,0),(1,1) satisfies your boundary conditions.
> And you can use whatever input the user's providing to
> determine as many intermediate control points as you want,
> and/or let the user explicitly specify them.
> Here's a short routine I use for bezier curves, where
> "point" is the obvious struct, and "binky(n,k)" is the
> binomial coefficient n!/k!(n-k)! ...
>
> /*===========================================================================
> * Function: bezier ( n, cp, t )
> * Purpose: returns x,y-coords of bezier curve
> * with n control points cp, at 0.<=t<=1.
> * --------------------------------------------------------------------------
> * Arguments: n (I) int n>=2, number of control points
> * cp (I) point* vector of control points, with
> * cp[0] the initial, and cp[n-1] the final
> * points of the curve, as usual.
> * t (I) double 0.<=t<=1.
> * Returns: (point) x,y-coords of bezier curve at t,
> * or -1.,-1. for argument error, or any error.
> * --------------------------------------------------------------------------
> * Notes: o
> *======================================================================== */
> /* --- entry point --- */
> point bezier ( int n, point *cp, double t ) {
> /* --- allocations and declarations --- */
> point bezt = { -1., -1. }; /* returned point, init for error */
> int k=0, binky(); /* binomial coefficient */
> /* --- check args --- */
> if ( n<2 || cp==NULL /* invalid control points */
> || t<0. || t>1. ) goto end_of_job; /* invalid parameter */
> /* --- explicit sum over control points --- */
> n--; /* sum from 0...n rather than 0...n-1 */
> bezt.x = bezt.y = 0.0; /* init sum */
> for ( k=0; k<=n; k++ ) {
> double coef=((double)binky(n,k))*pow(1.-t,(double)(n-k))*pow(t,(double)k);
> bezt.x += coef*cp[k].x;
> bezt.y += coef*cp[k].y;
> } /* --- end-of-for(k) --- */
> end_of_job:
> return ( bezt ); /* back to caller */
> } /* --- end-of-function bezier() --- */
>
> --
> John Forkosh ( mailto: j...@f.com where j=john and f=forkosh )
>
S-shaped means that it's got one point of inflection. So it increases slowly at first,
the steeply, then slowly. The curve starts off as concave (from the top) and then
switches to convex. Alternatively it starts off steeply, becomes shallow, and
then becomes steep again. So the curve starts off convex and becomes concave.

A cubic (?) Bezier is a good idea. That satisfies the boundary conditions. Whilst it's easy
to get the inflection point of a Bezier, the problem is going backwards, to create a Bezier
with an inflection point at the midpoint of the markers, and with extra constraints
to make it monotonic, and to make steepness relate to the markers in some intuitive
way.

Re: Anyone got a nice S-shaped function?

<747c0119-5fd4-4830-82ed-7ffb82ad18d5n@googlegroups.com>

 copy mid

https://www.novabbs.com/devel/article-flat.php?id=36796&group=comp.theory#36796

 copy link   Newsgroups: comp.theory
X-Received: by 2002:a05:622a:410:b0:33b:72b2:6038 with SMTP id n16-20020a05622a041000b0033b72b26038mr10438016qtx.627.1659805135326;
Sat, 06 Aug 2022 09:58:55 -0700 (PDT)
X-Received: by 2002:a81:6fc3:0:b0:323:6f8b:f169 with SMTP id
k186-20020a816fc3000000b003236f8bf169mr11037934ywc.494.1659805135165; Sat, 06
Aug 2022 09:58:55 -0700 (PDT)
Path: i2pn2.org!i2pn.org!usenet.blueworldhosting.com!feed1.usenet.blueworldhosting.com!peer01.iad!feed-me.highwinds-media.com!news.highwinds-media.com!news-out.google.com!nntp.google.com!postnews.google.com!google-groups.googlegroups.com!not-for-mail
Newsgroups: comp.theory
Date: Sat, 6 Aug 2022 09:58:54 -0700 (PDT)
In-Reply-To: <f9dc1b80-9215-44ed-802f-98922df6d265n@googlegroups.com>
Injection-Info: google-groups.googlegroups.com; posting-host=173.53.104.152; posting-account=X_pe-goAAACrVTtZeoCLt7hslVPY2-Uo
NNTP-Posting-Host: 173.53.104.152
References: <f9dc1b80-9215-44ed-802f-98922df6d265n@googlegroups.com>
User-Agent: G2/1.0
MIME-Version: 1.0
Message-ID: <747c0119-5fd4-4830-82ed-7ffb82ad18d5n@googlegroups.com>
Subject: Re: Anyone got a nice S-shaped function?
From: xlt....@gmail.com (B.H.)
Injection-Date: Sat, 06 Aug 2022 16:58:55 +0000
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
X-Received-Bytes: 2571
 by: B.H. - Sat, 6 Aug 2022 16:58 UTC

On Friday, August 5, 2022 at 6:59:03 AM UTC-4, malcolm.ar...@gmail.com wrote:
> I'm looking for a nice S-shaped function which is zero at x= 00, 1 at x = 1, and is montonically increasing.
> I want it to be possible to make it very close to the sum of absolute derivatives of the cosine curve (this is the curve you obtain by placing marks at even spaces on the edge of a semicircle, then take the x-coordinates), and I want it to respond intuitively to two markers I'll give to the user. The mid point of the two markers should be the inflection point of the curve, the distance from the marker to the x=0 or x =1 point represents the steepness of the curve.
>
> (What I want it for is to allow the user to specify curved surfaces without having a full 3D drawing system. Most of those surfaces will be semicircular because he'll be drawing a can or a bottle or something similar. Hence the cosine-based curve will be the default. But I want him to have a bit of control
> to tweak the surface).

Why don't you just do something along the lines of:

f(x) = (x-a)^(2*k+1) - b

....where k is a constant natural number, and a, b, help you manipulate the "origin" of the curve?

If you want to stretch it out vertically or horizontally, just multiply (f(x)+b) or (x-a) by a constant.

-Philip White (philipjwhite@yahoo.com)

Re: Anyone got a nice S-shaped function?

<56d3af49-49c9-49a0-b11f-6361244543aan@googlegroups.com>

 copy mid

https://www.novabbs.com/devel/article-flat.php?id=36797&group=comp.theory#36797

 copy link   Newsgroups: comp.theory
X-Received: by 2002:a05:622a:1448:b0:341:e523:fd18 with SMTP id v8-20020a05622a144800b00341e523fd18mr10009017qtx.338.1659805243759;
Sat, 06 Aug 2022 10:00:43 -0700 (PDT)
X-Received: by 2002:a25:401:0:b0:677:6c32:b454 with SMTP id
1-20020a250401000000b006776c32b454mr9744488ybe.52.1659805243458; Sat, 06 Aug
2022 10:00:43 -0700 (PDT)
Path: i2pn2.org!i2pn.org!usenet.blueworldhosting.com!feed1.usenet.blueworldhosting.com!peer01.iad!feed-me.highwinds-media.com!news.highwinds-media.com!news-out.google.com!nntp.google.com!postnews.google.com!google-groups.googlegroups.com!not-for-mail
Newsgroups: comp.theory
Date: Sat, 6 Aug 2022 10:00:43 -0700 (PDT)
In-Reply-To: <747c0119-5fd4-4830-82ed-7ffb82ad18d5n@googlegroups.com>
Injection-Info: google-groups.googlegroups.com; posting-host=173.53.104.152; posting-account=X_pe-goAAACrVTtZeoCLt7hslVPY2-Uo
NNTP-Posting-Host: 173.53.104.152
References: <f9dc1b80-9215-44ed-802f-98922df6d265n@googlegroups.com> <747c0119-5fd4-4830-82ed-7ffb82ad18d5n@googlegroups.com>
User-Agent: G2/1.0
MIME-Version: 1.0
Message-ID: <56d3af49-49c9-49a0-b11f-6361244543aan@googlegroups.com>
Subject: Re: Anyone got a nice S-shaped function?
From: xlt....@gmail.com (B.H.)
Injection-Date: Sat, 06 Aug 2022 17:00:43 +0000
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
X-Received-Bytes: 2877
 by: B.H. - Sat, 6 Aug 2022 17:00 UTC

On Saturday, August 6, 2022 at 12:58:56 PM UTC-4, B.H. wrote:
> On Friday, August 5, 2022 at 6:59:03 AM UTC-4, malcolm.ar...@gmail.com wrote:
> > I'm looking for a nice S-shaped function which is zero at x= 00, 1 at x = 1, and is montonically increasing.
> > I want it to be possible to make it very close to the sum of absolute derivatives of the cosine curve (this is the curve you obtain by placing marks at even spaces on the edge of a semicircle, then take the x-coordinates), and I want it to respond intuitively to two markers I'll give to the user.. The mid point of the two markers should be the inflection point of the curve, the distance from the marker to the x=0 or x =1 point represents the steepness of the curve.
> >
> > (What I want it for is to allow the user to specify curved surfaces without having a full 3D drawing system. Most of those surfaces will be semicircular because he'll be drawing a can or a bottle or something similar. Hence the cosine-based curve will be the default. But I want him to have a bit of control
> > to tweak the surface).
> Why don't you just do something along the lines of:
>
> f(x) = (x-a)^(2*k+1) - b
>
> ...where k is a constant natural number, and a, b, help you manipulate the "origin" of the curve?
>
> If you want to stretch it out vertically or horizontally, just multiply (f(x)+b) or (x-a) by a constant.
>
> -Philip White (philip...@yahoo.com)

Also, if you want it to look more like a cosine curve, examine the Taylor series expansion based on f(x) = cos (x) .

-Philip White

Re: Anyone got a nice S-shaped function?

<48925441-14eb-4423-9a16-b50234cdf797n@googlegroups.com>

 copy mid

https://www.novabbs.com/devel/article-flat.php?id=36798&group=comp.theory#36798

 copy link   Newsgroups: comp.theory
X-Received: by 2002:a05:6214:2a82:b0:476:b707:e1c4 with SMTP id jr2-20020a0562142a8200b00476b707e1c4mr10072882qvb.99.1659805310776;
Sat, 06 Aug 2022 10:01:50 -0700 (PDT)
X-Received: by 2002:a81:8413:0:b0:322:d661:a785 with SMTP id
u19-20020a818413000000b00322d661a785mr10861598ywf.16.1659805310544; Sat, 06
Aug 2022 10:01:50 -0700 (PDT)
Path: i2pn2.org!i2pn.org!usenet.blueworldhosting.com!feed1.usenet.blueworldhosting.com!peer01.iad!feed-me.highwinds-media.com!news.highwinds-media.com!news-out.google.com!nntp.google.com!postnews.google.com!google-groups.googlegroups.com!not-for-mail
Newsgroups: comp.theory
Date: Sat, 6 Aug 2022 10:01:50 -0700 (PDT)
In-Reply-To: <56d3af49-49c9-49a0-b11f-6361244543aan@googlegroups.com>
Injection-Info: google-groups.googlegroups.com; posting-host=173.53.104.152; posting-account=X_pe-goAAACrVTtZeoCLt7hslVPY2-Uo
NNTP-Posting-Host: 173.53.104.152
References: <f9dc1b80-9215-44ed-802f-98922df6d265n@googlegroups.com>
<747c0119-5fd4-4830-82ed-7ffb82ad18d5n@googlegroups.com> <56d3af49-49c9-49a0-b11f-6361244543aan@googlegroups.com>
User-Agent: G2/1.0
MIME-Version: 1.0
Message-ID: <48925441-14eb-4423-9a16-b50234cdf797n@googlegroups.com>
Subject: Re: Anyone got a nice S-shaped function?
From: xlt....@gmail.com (B.H.)
Injection-Date: Sat, 06 Aug 2022 17:01:50 +0000
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
X-Received-Bytes: 3154
 by: B.H. - Sat, 6 Aug 2022 17:01 UTC

On Saturday, August 6, 2022 at 1:00:45 PM UTC-4, B.H. wrote:
> On Saturday, August 6, 2022 at 12:58:56 PM UTC-4, B.H. wrote:
> > On Friday, August 5, 2022 at 6:59:03 AM UTC-4, malcolm.ar...@gmail.com wrote:
> > > I'm looking for a nice S-shaped function which is zero at x= 00, 1 at x = 1, and is montonically increasing.
> > > I want it to be possible to make it very close to the sum of absolute derivatives of the cosine curve (this is the curve you obtain by placing marks at even spaces on the edge of a semicircle, then take the x-coordinates), and I want it to respond intuitively to two markers I'll give to the user. The mid point of the two markers should be the inflection point of the curve, the distance from the marker to the x=0 or x =1 point represents the steepness of the curve.
> > >
> > > (What I want it for is to allow the user to specify curved surfaces without having a full 3D drawing system. Most of those surfaces will be semicircular because he'll be drawing a can or a bottle or something similar. Hence the cosine-based curve will be the default. But I want him to have a bit of control
> > > to tweak the surface).
> > Why don't you just do something along the lines of:
> >
> > f(x) = (x-a)^(2*k+1) - b
> >
> > ...where k is a constant natural number, and a, b, help you manipulate the "origin" of the curve?
> >
> > If you want to stretch it out vertically or horizontally, just multiply (f(x)+b) or (x-a) by a constant.
> >
> > -Philip White (philip...@yahoo.com)
> Also, if you want it to look more like a cosine curve, examine the Taylor series expansion based on f(x) = cos (x) .
>
> -Philip White

Btw, check out desmos.com if you want to experiment with a good graphing calculator.

-Philip White

Re: Anyone got a nice S-shaped function?

<87y1w1qcmm.fsf@nosuchdomain.example.com>

 copy mid

https://www.novabbs.com/devel/article-flat.php?id=36801&group=comp.theory#36801

 copy link   Newsgroups: comp.theory
Path: i2pn2.org!i2pn.org!eternal-september.org!reader01.eternal-september.org!.POSTED!not-for-mail
From: Keith.S....@gmail.com (Keith Thompson)
Newsgroups: comp.theory
Subject: Re: Anyone got a nice S-shaped function?
Date: Sat, 06 Aug 2022 14:35:13 -0700
Organization: None to speak of
Lines: 11
Message-ID: <87y1w1qcmm.fsf@nosuchdomain.example.com>
References: <f9dc1b80-9215-44ed-802f-98922df6d265n@googlegroups.com>
MIME-Version: 1.0
Content-Type: text/plain
Injection-Info: reader01.eternal-september.org; posting-host="de883fb70249bf3e4c71e17869dbfd2d";
logging-data="96007"; mail-complaints-to="abuse@eternal-september.org"; posting-account="U2FsdGVkX1/cOuLSEr8nNSh7DvcJfb6b"
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/27.2 (gnu/linux)
Cancel-Lock: sha1:1dJkNtsXyE0/zcgwVqPqB/CamL4=
sha1:HSbyxZm50u3Q+EUHScsJTiyBvXw=
 by: Keith Thompson - Sat, 6 Aug 2022 21:35 UTC

Malcolm McLean <malcolm.arthur.mclean@gmail.com> writes:
> I'm looking for a nice S-shaped function which is zero at x= 00, 1 at
> x = 1, and is montonically increasing.
[...]

4*arctan(x)/pi ?

--
Keith Thompson (The_Other_Keith) Keith.S.Thompson+u@gmail.com
Working, but not speaking, for Philips
void Void(void) { Void(); } /* The recursive call of the void */

1
server_pubkey.txt

rocksolid light 0.9.7
clearnet tor