Re: Supported Floating Point Types

Athanasios Iliopoulos via ublas <[email protected]> Fri, 18 Jan 2019 15:56:19 +0000
Newsgroups gmane.comp.lib.boost.ublas
Message-ID <SN6PR11MB33276A25754997C4CA27E56FD49C0@SN6PR11MB3327.namprd11.prod.outlook.com>
--===============8320133462389458431==
Content-Language: en-US
Content-Type: multipart/alternative;
	boundary="_000_SN6PR11MB33276A25754997C4CA27E56FD49C0SN6PR11MB3327namp_"

--_000_SN6PR11MB33276A25754997C4CA27E56FD49C0SN6PR11MB3327namp_
Content-Type: text/plain; charset="Windows-1252"
Content-Transfer-Encoding: quoted-printable

Any POD like numeric type is supported in uBlas. As long as the regular ope=
rators (+,-,*) are defined for that type, and it has regular construction s=
emantics. You can even use it with boost multiprecision to achieve arbitrar=
y precision arithmetic.

Something that may be an issue is the behavior of these types with streams.=
 You may need to overload operator<< for these to behave properly.

-N

On 1/18/19 9:32 AM, Robert Firl via ublas wrote:
Hello Everyone,

Note that I have not reviewed all of the documentation.  I have looked at t=
he Overview page.  Plus, I have performed uBLAS include directory searches =
on the phrases:

long double
double
float128
float80
float64

What I am trying to determine is the highest precision floating point type =
supported by uBLAS.  It looks to be long double, hopefully as an 80-bit rep=
resentation.  Is this correct?

If not, if a compiler supports 128-bit floating point, does uBLAS through t=
he type_traits, value_type use support the boost::float128_t type?

Here at my employer, there is a series of algorithmic tests we have to do o=
n an implementation with double.  The current implementation may be losing =
precision.  The tests need higher precision to prove whether this hypothesi=
s is correct. An 80-bit long double is enough.  However, if the boost::floa=
t128_t is fully supported in uBLAS, then it is already known how to fully e=
nable the 128-bit floating point support. Thank you all for your help.

Sincerely,

Robert

PS The Gmane newsgroup auto authorization e-mail reply address has an issue=
.  Three times now, I have replied to the authorization message.  Within a =
minute of clicking send, I receive a blaine.gmane.org rejected message that=
 the "authorization code"@auth.gmane.org has status 550 unknown user.  I as=
sume this indicates a Gmane issue.  Is that correct?




_______________________________________________
ublas mailing list
[email protected]<mailto:[email protected]>
https://lists.boost.org/mailman/listinfo.cgi/ublas
Sent to: [email protected]<mailto:athanasios.iliopo=
[email protected]>


--_000_SN6PR11MB33276A25754997C4CA27E56FD49C0SN6PR11MB3327namp_
Content-Type: text/html; charset="Windows-1252"
Content-ID: <[email protected]>
Content-Transfer-Encoding: quoted-printable

<html>
<head>
<meta http-equiv=3D"Content-Type" content=3D"text/html; charset=3DWindows-1=
252">
</head>
<body text=3D"#000000" bgcolor=3D"#FFFFFF">
<p>Any POD like numeric type is supported in uBlas. As long as the regular =
operators (&#43;,-,*) are defined for that type, and it has regular constru=
ction semantics. You can even use it with boost multiprecision to achieve a=
rbitrary precision arithmetic.</p>
<p>Something that may be an issue is the behavior of these types with strea=
ms. You may need to overload operator&lt;&lt; for these to behave properly.=
<br>
</p>
<p>-N<br>
</p>
<div class=3D"moz-cite-prefix">On 1/18/19 9:32 AM, Robert Firl via ublas wr=
ote:<br>
</div>
<blockquote type=3D"cite" cite=3D"mid:6ca1e7a419b64eb6a5742f0a32dbdedc@bre.=
com">
<meta name=3D"Generator" content=3D"Microsoft Word 15 (filtered
        medium)">
<style><!--
/* Font Definitions */
@font-face
	{font-family:"Cambria Math";
	panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
	{font-family:Calibri;
	panose-1:2 15 5 2 2 2 4 3 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
	{margin:0in;
	margin-bottom:.0001pt;
	font-size:11.0pt;
	font-family:"Calibri",sans-serif;}
a:link, span.MsoHyperlink
	{mso-style-priority:99;
	color:#0563C1;
	text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
	{mso-style-priority:99;
	color:#954F72;
	text-decoration:underline;}
span.EmailStyle17
	{mso-style-type:personal-compose;
	font-family:"Calibri",sans-serif;
	color:windowtext;}
.MsoChpDefault
	{mso-style-type:export-only;
	font-family:"Calibri",sans-serif;}
@page WordSection1
	{size:8.5in 11.0in;
	margin:1.0in 1.0in 1.0in 1.0in;}
div.WordSection1
	{page:WordSection1;}
--></style><!--[if gte mso 9]><xml>
<o:shapedefaults v:ext=3D"edit" spidmax=3D"1026" />
</xml><![endif]--><!--[if gte mso 9]><xml>
<o:shapelayout v:ext=3D"edit">
<o:idmap v:ext=3D"edit" data=3D"1" />
</o:shapelayout></xml><![endif]-->
<div class=3D"WordSection1">
<p class=3D"MsoNormal">Hello Everyone,<o:p></o:p></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">Note that I have not reviewed all of the documentati=
on.&nbsp; I have looked at the Overview page.&nbsp; Plus, I have performed =
uBLAS include directory searches on the phrases:<o:p></o:p></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">long double<o:p></o:p></p>
<p class=3D"MsoNormal">double<o:p></o:p></p>
<p class=3D"MsoNormal">float128<o:p></o:p></p>
<p class=3D"MsoNormal">float80<o:p></o:p></p>
<p class=3D"MsoNormal">float64<o:p></o:p></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">What I am trying to determine is the highest precisi=
on floating point type supported by uBLAS.&nbsp; It looks to be long double=
, hopefully as an 80-bit representation.&nbsp; Is this correct?<o:p></o:p><=
/p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">If not, if a compiler supports 128-bit floating poin=
t, does uBLAS through the type_traits, value_type use support the boost::fl=
oat128_t type?<o:p></o:p></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">Here at my employer, there is a series of algorithmi=
c tests we have to do on an implementation with double.&nbsp; The current i=
mplementation may be losing precision.&nbsp; The tests need higher precisio=
n to prove whether this hypothesis is correct.
 An 80-bit long double is enough.&nbsp; However, if the boost::float128_t i=
s fully supported in uBLAS, then it is already known how to fully enable th=
e 128-bit floating point support. Thank you all for your help.<o:p></o:p></=
p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">Sincerely,<o:p></o:p></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">Robert <o:p></o:p></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
<p class=3D"MsoNormal">PS The Gmane newsgroup auto authorization e-mail rep=
ly address has an issue.&nbsp; Three times now, I have replied to the autho=
rization message.&nbsp; Within a minute of clicking send, I receive a blain=
e.gmane.org rejected message that the &quot;authorization
 code&quot;@auth.gmane.org has status 550 unknown user.&nbsp; I assume this=
 indicates a Gmane issue.&nbsp; Is that correct?<o:p></o:p></p>
<p class=3D"MsoNormal"><o:p>&nbsp;</o:p></p>
</div>
<br>
<fieldset class=3D"mimeAttachmentHeader"></fieldset>
<pre class=3D"moz-quote-pre" wrap=3D"">____________________________________=
___________
ublas mailing list
<a class=3D"moz-txt-link-abbreviated" href=3D"mailto:[email protected]"=
>[email protected]</a>
<a class=3D"moz-txt-link-freetext" href=3D"https://lists.boost.org/mailman/=
listinfo.cgi/ublas">https://lists.boost.org/mailman/listinfo.cgi/ublas</a>
Sent to: <a class=3D"moz-txt-link-abbreviated" href=3D"mailto:athanasios.il=
[email protected]">[email protected]</a>
</pre>
</blockquote>
</body>
</html>

--_000_SN6PR11MB33276A25754997C4CA27E56FD49C0SN6PR11MB3327namp_--

--===============8320133462389458431==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline