Imaginary i in Latex
Thierry Bouche
> Isn't an ordinary i normally used?
there are good reason for which it should be roman (as the e in
exponential, and even the d of differentials--same kind of reasons for
which sin or cos are in roman). But something like 90 % of the
printers aren't aware of these good reasons...
Bo Thide'
>
> Hi, how can I write the imaginary i in LaTeX, what's the standart for it?
>
> Best regards,
In Europe, it is considered good (math) typography to set the imaginary
unit as an upright "i", e.g.,
\renewcommand{\i}{{\mathrm{i}}
to discriminate it from the usual math italic "i" which really denotes
an integer index (sub-/superscript, summation index or whatever).
Check a math publication printed by a well-known quality European publishing
house and you will see what I mean.
US math typography is a bit more sloppy/forgiving and some US publishers
really insist that the imaginary unit and the interger index i should be
identical, even though it may lead to confusions...
Bo
--
^ Bo Thidé, PhD, Professor of Space Physics----www.wavegroup.irfu.se/~bt
|I| Swedish Institute of Space Physics, P. O. Box 537, SE-752 21 Uppsala
|R| Office: Ångström Laboratory, Room 14144, Lägerhyddsvägen 1, Uppsala
/|F|\ Phone: 018-4715914. Fax: 018-4715905. Mobile/Cell Phone: 0705-613670
~~U~~ E-Mail: mailto:b...@irfu.se Mobile/Cell e-mail: mailto:cal...@irfu.se
dlue...@my-deja.com
b...@irfu.se wrote:
> "Paulo J. Matos aka PDestroy" wrote:
> >
> > Hi, how can I write the imaginary i in LaTeX, what's the standart
for it?
>
> In Europe, it is considered good (math) typography to set the
imaginary
> unit as an upright "i", e.g.,
>
> \renewcommand{\i}{{\mathrm{i}}
>
> to discriminate it from the usual math italic "i" which really denotes
> an integer index (sub-/superscript, summation index or whatever).
> Check a math publication printed by a well-known quality European
publishing
> house and you will see what I mean.
Pulling out my copies of articles from Springer-Verlag math journals, I
do not find one upright i for the imaginary unit. Does this mean
Springer is not a "quality European Publishing house"?
>
> US math typography is a bit more sloppy/forgiving and some US
publishers
> really insist that the imaginary unit and the interger index i should
be
> identical, even though it may lead to confusions...
The only confusion that can come is in using both senses of i in the
same place. A sensible person shouldn't do that with any symbol, and
should avoid even using the same character in different fonts, such as
$\sum_{i=1}^\infty a_i + b_i\mathrm{i}$. Something like
$\sum_{k=1}^\infty a_k + b_k i$ would be a ten times better (well two
times anyway).
Since Complex Variables is my specialty, I have read and written tens of
thousands of math italic imaginary units and have yet to be confused (at
least not by that). And the same is true of e for the Euler constant and
d for a differential. The context generally makes all very clear: e and
d are (or should be) seldom used for more than one purpose in any given
article.
--
Dan Luecking
University of Arkansas
Sent via Deja.com
http://www.deja.com/
Thomas Ruedas
>d are (or should be) seldom used for more than one purpose in any given
>article.
considered good style to use an upright d for the differential because
it is an operator and not a variable or constant, in contrast to e or i.
Admittedly, it seems that many people don't care; I have seen many
otherwise nicely typeset books where the operator was typeset in math
italics.
--
------------------------------------------------------------------------
Thomas Ruedas
Institute of Meteorology and Geophysics, J.W.Goethe University Frankfurt
e-mail: rue...@geophysik.uni-frankfurt.de
http://www.geophysik.uni-frankfurt.de/~ruedas/
------------------------------------------------------------------------
Xiaotian Sun
>>d for a differential. The context generally makes all very clear: e and
>>d are (or should be) seldom used for more than one purpose in any given
>>article.
> In principle I agree, although IMHO d is a special case. AFAIK it is
> considered good style to use an upright d for the differential because
> it is an operator and not a variable or constant, in contrast to e or i.
> Admittedly, it seems that many people don't care; I have seen many
> otherwise nicely typeset books where the operator was typeset in math
> italics.
actually I was wondering why LaTeX or amsmath package didn't
difine a command for the differential operator d during the
first days I learned LaTeX. I had to define my own one for it.
Xiaotian
Ilya Zakharevich
<kna...@springer.de>],
who wrote in article <3A81013D...@springer.de>:
> Different sciences (and publishers) disagree here. Mathematicians and
> Springer-Verlag prefer a math italic i, phyisicists and some other
> publishers prefer the math roman i.
I think using upright letters for constants is a "French style". The
alternative styles are to use \sqrt{-1} (as normal in motives theory),
or do not use i for variables. Use letters starting with j as indices
of summation.
The french style does not help when you write on blackboard, so not
all the people buy it. But on paper it is quite OK.
Ilya
dlue...@my-deja.com
Thomas Ruedas <rue...@geophysik.uni-frankfurt.de> wrote:
> >d for a differential. The context generally makes all very clear: e
and
> >d are (or should be) seldom used for more than one purpose in any
given
> >article.
> In principle I agree, although IMHO d is a special case. AFAIK it is
> considered good style to use an upright d for the differential because
> it is an operator and not a variable or constant, in contrast to e or
i.
In principle, I agree that d is an operator. However it differs from
many other operators both historically and aesthetically.
Historically, items like dx and dy came to be treated as a single entity
and it clashes to see the different parts of it typeset differently.
Also, in the combination \int f(x) \,dx, the dx serves mostly to
terminate the integral and to indicate the variable (in my handwritten
note I write \int_x f(x). where the dot and the subscript serve these
two purposes).
Aesthetically, single-letter (or even single-symbol operators, if they
are "small") seem to benefit from being treated differently than
"log-like" operators or "big operators (like \sum). I once wrote a paper
in which \bar\partial appeared as an operator many times. I tried
\def\dbar{\mathop{\bar\partial}\nolimits} and decided I didn't like the
spacing that really looks so nice with other TeX \mathop-s. I feel the
same about d, \mathrm{d}x just looks butt-ugly to me, and still worse is
\mathop{\kern0pt\mathrm d}\nolimits x.
> Admittedly, it seems that many people don't care; I have seen many
> otherwise nicely typeset books where the operator was typeset in math
> italics.
I care, I just disagree on what works best.
Thierry Bouche
> Since Complex Variables is my specialty,
same for me. The choice of roman e, i, d is not meant to avoid
confusion (although it's a feature), it's meant for better semantics
and more coherence in the notation. Now these are principles, and
these principles obviously are not applied, and their non application
doesn't do much harm.
BTW Springer appears to mostly publish author's unverified _copies_ as
long as typography is concerned, so you can't expect any homogeneity
in notation from them.
regarding \mathrm{d}x, yes it's typically ugly when you're not used
to, and no it should not be spaced as a \mathop. When you're used to,
it looks as normal as tr A or anything alike.
--
Thierry Bouche
__
« Ils vivent pour vivre, et nous, hélas ! nous vivons pour savoir. »
Charles Baudelaire, Paris.
Timothy Murphy
>I think using upright letters for constants is a "French style".
I just looked through a few volumes of Bourbaki,
and he (they) seem to use math italic i, e and d like everyone else.
I would find it rather idiosyncratic if a math publisher
used a special symbol for i or e.
It might be logical, but it just ain't done.
--
Timothy Murphy
e-mail: t...@maths.tcd.ie
tel: 086-233 6090
s-mail: School of Mathematics, Trinity College, Dublin 2, Ireland
Timothy Murphy
>regarding \mathrm{d}x, yes it's typically ugly when you're not used
>to, and no it should not be spaced as a \mathop. When you're used to,
>it looks as normal as tr A or anything alike.
There is an obvious reason for printing "tr" in roman --
to avoid confusion with t \times r.
Incidentally, which French math books use the style you recommend?
I've just glanced through a few (mainly Hermann) and I can't find any.
J%org Knappen
>
> Incidentally, which French math books use the style you recommend?
> I've just glanced through a few (mainly Hermann) and I can't find any.
About a decade ago, I collected the differences in style of several
physics journals. ISO style with upright d, e, and i was best followed
by the italian journal Il Nouvo Cimento, french style was *very*
different
(with all capital letters upright), If I remeber right the journal was
Journal de Physique.
No need to mention that the journals cited above do not exist any
longer as independent journals, but have merged with Zeitschrift für
Physik into a joint venture of serveral publishers under the journal
label European Journal of Physics.
If have written a paper for the EuroTeX conference at Prague with
all of this and I had a protype style file for LaTeX2.09/NFSS1
implementing switches for different math styles. It is still not
ported to LaTeX2e and remains unpublished (though it should still
exist on my disk space somewhere).
--J"org Knappen
Thierry Bouche
> Thierry Bouche <Thierry...@ujf-grenoble.fr> writes:
>
> >regarding \mathrm{d}x, yes it's typically ugly when you're not used
> >to, and no it should not be spaced as a \mathop. When you're used to,
> >it looks as normal as tr A or anything alike.
>
> There is an obvious reason for printing "tr" in roman --
> to avoid confusion with t \times r.
the same reason for which dx should read as d \times x (which it
happens to mean sometimes...)
> Incidentally, which French math books use the style you recommend?
> I've just glanced through a few (mainly Hermann) and I can't find any.
I never said this was french style. I have seen french books using
upright e, i, d (more often i, e but not d); it seems to be more
frequent in physics or natural sciences than in maths.
As Jörg said, french style is more recognisable by its upright
capitals and greek. I realised only recently that this had to do with
the existence of `I' in english.
To be consistent with the above discussion, there should be 2 \pi: one
upright for the universal constant, and one for other pi's (typically:
projections). in the same way as there are `textual mu's' (typically
upright) for micro-, and math mu.
--
Thierry Bouche
Timothy Murphy
>To be consistent with the above discussion, there should be 2 \pi: one
>upright for the universal constant, and one for other pi's (typically:
>projections). in the same way as there are `textual mu's' (typically
>upright) for micro-, and math mu.
And one would have to use different 2's for a^2 in a squared
and a^2 as in an entry in the vector a^i.
It is obvious (to me) that if you tried to distinguish
all such cases by using different symbols one would go (or already be) mad.
So one has to reach a reasonable compromise,
which I take to be that different symbols should be used
if there is a measurable chance of mis-interpretation otherwise.
For example, I have never been confused by thinking that "dx"
was a product when it was really a differential.
The malign influence of XML is all too prevalent ...
Michael J Downes
> actually I was wondering why LaTeX or amsmath package didn't
> difine a command for the differential operator d during the
> first days I learned LaTeX. I had to define my own one for it.
That is one of the things we have planned to include in an auxiliary
package, but probably not directly in amsmath because of the forward
compatibility problem (e.g., if author A with version 3.x of amsmath
sends a document using \diffd to coauthor B who is still using version
2.x, author B is not going to like the "Undefined control sequence"
error).
Regards, Michael Downes
tech-s...@ams.org