3

I am writing my thesis in physics.

I am not a Tex expert so in the beginning I just added convenient packages that were suggested here on tex.stackexchange.com as I went along. Unfortunately, after some time, I noticed that some of these packages conflict with each other and some even produce wrong results.

I think there should be a list of packages that are NOT recommended for ignorant newbies like me in order to save time. I can see how this may be a highly subjective issue, but I did not find any such consolidated discussion elsewhere, so I think it is worth a post.

Here is a list of packages that I am/was considering in alphabetical order.

aas_macros
acronym
amsmath
amssymb
appendix
booktabs
braket
breqn
calc
caption
cite
cleveref
commath
datetime2
enumitem
etoolbox
expl3
fancyhdr
feynmp
geometry
graphicx
hepnames
hyperref
latexsym
mathtools
mhchem
microtype
nag
physics
rotating
siunitx
subcaption
tabularx
tensor
xfrac
xifthen
xparse
xspace

Two packages that I already know have issues: breqn, commath.

Is there any package here that someone with experience can recommend against?

mich
  • 63
  • 2
  • 6
  • 1
    Alphabetical is not very useful. Break them into categories; e.g., tables, lists, graphics, bibliography, fonts, etc. (Note this might mean taking the time to look at package documentation!) Also, don't just dump old preambles in new documents. Take some time and add things as needed. Eventually, you will develop your core set of indispensable packages. – jon Jul 12 '15 at 02:30
  • I strongly recommend not using all those packages which you do not need i.e. are not using. Note that you are certainly loading some packages unnecessarily because they will be loaded by other packages anyway. Note also that whether a package is a good idea or not depends partly on your document class. Finally, this seems to have very little to do with physics, to be honest. – cfr Jul 12 '15 at 02:35
  • Related: a question of mine about discipline-specific recommendations and an answer concerning a more general question. – cfr Jul 12 '15 at 02:38
  • 1
    And what is the class used? Potential problems might arise from loading packages with certain classes. – Gonzalo Medina Jul 12 '15 at 02:55
  • 1
    I think this is opinion-based question. For me, the best is to start from a minimal document (just the class, may be amsmath and geometry packages) and add packages as needed, this help you learning and show you any conflict – touhami Jul 12 '15 at 06:35
  • 1
    the question doesn't make sense as it stands, if you need the functionality of a package then you need that functionality so either that package or another one that offers equivalent functionality. So you need to ask a question like I need xxx feature should I use package X or Y. Also the packages you list are all at different levels expl3 for example is no use at all in a document but if you need siunitx then it will load expl3 as it uses it so asking separately whether to use siunitx and expl3 can not have a useful answer. – David Carlisle Jul 12 '15 at 09:36
  • When I was using postscript packages and switched to PdfLaTeX many things broke. Since full-blown postscript isn't compatible with pdf I decided to look for different solutions. Initially I was happy using MetaPost but now I use TikZ for all my diagrams. –  Jul 12 '15 at 10:54

1 Answers1

3

The basic advice should be to start with a documentclass that is closest to your needs/or standard in your area/fits with a journal you are intending to submit to. After that only add packages if you find you need them for functionality not provide by the class.

A standard class in physics is revtex or rather revtex4-1. The documentation contains an author guide, that discusses features and suitable packages to load to solve certain common problems.

Andrew Swann
  • 95,762