From owner-chemistry@ccl.net Thu Aug 16 21:19:00 2012
From: "Mark Zottola mzottola~!~gmail.com" <owner-chemistry _ server.ccl.net>
To: CCL
Subject: CCL: Dipole Moment Question
Message-Id: <-47372-120816210539-4094-Hi1Otpig7SC3u2MpzY/L3Q _ server.ccl.net>
X-Original-From: Mark Zottola <mzottola%x%gmail.com>
Content-Type: multipart/alternative; boundary=e89a8f83ac518b18c504c76bc1d7
Date: Thu, 16 Aug 2012 21:05:29 -0400
MIME-Version: 1.0


Sent to CCL by: Mark Zottola [mzottola(0)gmail.com]
--e89a8f83ac518b18c504c76bc1d7
Content-Type: text/plain; charset=ISO-8859-1

I recently encountered a phenomenon that has me a bit baffled.

For fun, I was plotting the dipole moment of a molecule as a function of
the bond length (compression and expansion).  I obtained this data from a
set of relaxed potential energy scans.  It turned out for a range of
approximately 1.1 Angstroms of deformation the dipole moment varied as a
cubic function of the deformed bond length.  The interesting part of that
correlation was the goodness of fit was exactly 1.  In other words for a 12
point data set, the fit of the cubic to the data was perfect.

I am not overfitting the data and redoing the calculations at a different
level of theory and different basis set gave another perfectly fitting
cubic function.  I have reread the basic physics of dipole moment and see
no reason why I should see a such a perfect fit.

If I am missing something obvious - can I get a pointer to the appropriate
information?  Thanks.

--e89a8f83ac518b18c504c76bc1d7
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

<div>I recently encountered a phenomenon that has me a bit baffled.</div><d=
iv><br></div><div>For fun, I was plotting the dipole moment of a molecule a=
s a function of the bond length (compression and expansion). =A0I obtained =
this data from a set of relaxed potential energy scans. =A0It turned out fo=
r a range of approximately 1.1 Angstroms of deformation the dipole moment v=
aried as a cubic function of the deformed bond length. =A0The interesting p=
art of that correlation was the goodness of fit was exactly 1. =A0In other =
words for a 12 point data set, the fit of the cubic to the data was perfect=
.</div>
<div><br></div><div>I am not overfitting the data and redoing the calculati=
ons at a different level of theory and different basis set gave another per=
fectly fitting cubic function. =A0I have reread the basic physics of dipole=
 moment and see no reason why I should see a such a perfect fit. =A0</div>
<div><br></div><div>If I am missing something obvious - can I get a pointer=
 to the appropriate information? =A0Thanks.</div>

--e89a8f83ac518b18c504c76bc1d7--