Graphics, Figures & Tables ⇒ siunitx converstion of units

[LKB]

Joseph,
I am using Lyx as my front end but I dont think this make any difference. Testing following code

Code: Select all

\documentclass[oneside,british,oldfontcommands]{memoir}
\usepackage{babel}
\OnehalfSpacing
\makeatletter
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% User specified LaTeX commands.
\sisetup{per-mode=fraction, tight-spacing = true , fraction-function = \tfrac , quotient-mode = fraction}%
\sisetup{inter-unit-product = \ensuremath { { } \cdot { } } , exponent-product = \cdot }%
\sisetup{input-product=x , output-quotient =  \ensuremath { { } \times{}}} %for 1x2x3
\sisetup{group-minimum-digits = 4} %start grouping from 4 digits, in 3 no groups
\sisetup{range-units = single} %2-3C not 2C-3C
\sisetup{separate-uncertainty=true} %2+-1 not 2(1)
%\sisetup{scientific-notation = engineering }  % exp in multiple of 3
\sisetup{prefixes-as-symbols=false} %10^-9 ect ect
%\sisetup{zero-decimal-to-integer, round-mode = places,round-precision = 3}
%\DeclareSIUnit[scientific-notation = engineering, prefixes-as-symbols=false]
%  {\second}{s}
\DeclareSIUnit[scientific-notation = engineering, prefixes-as-symbols=false]
  {\ns}{\nano\second}

\makeatother

\begin{document}
\begin{center}
Where short term stability is $\SI{10(1)}{\nano\second}$ as well as $\SI{10(1)}{\ns}$, but our network edge is $\SI{10}{\km}$.
\par\end{center}
\end{document}

Show no display difference between $\SI{10(1)}{\nano\second}$ and $\SI{10(1)}{\ns}$ despite different settings

[cgnieder]

I can confirm LKB's observation. This MWE

Code: Select all

\documentclass{article}
\usepackage{siunitx}

\sisetup{
  scientific-notation=false,
  prefixes-as-symbols=true
}

\DeclareSIUnit[scientific-notation=engineering,prefixes-as-symbols=false]{\second}{s}
\DeclareSIUnit[scientific-notation=engineering,prefixes-as-symbols=false]{\ns}{\nano\second}

\begin{document}

\SI{10e2}{\nano\second}
\SI{10e2}{\kilo\metre}

\end{document}

on TL 2012 with »siunitx« v2.5g gives

siunitx.png (2.25 KiB) Viewed 8523 times

[LKB]

Is there any more news on this one? Does it mean that unit specific definition cannot be made?

[josephwright]

Per unit settings work for that unit command alone: settings for \second will not apply to \ns, for example.

[LKB]

If you look at Klemens example it does not work. Unit only definitions are not applied in the document in comparison to general command using \sisetup.

I think that way around it would be

Code: Select all

\newcommand*{\FlexUnits}[2]{\SI[scientific-notation=engineering,prefixes-as-symbols=false]{#1}{#2}}

but is not as nice nor flexible as using siunitx only

[josephwright]

It does work, but as I say only for exactly the units set up

Code: Select all

    \documentclass{article}
    \usepackage{siunitx}

    \sisetup{
      scientific-notation=false,
      prefixes-as-symbols=true
    }

    \DeclareSIUnit[scientific-notation=engineering,prefixes-as-symbols=false]{\second}{s}
    \DeclareSIUnit[scientific-notation=engineering,prefixes-as-symbols=false]{\nanosecond}{\nano\second}
    \DeclareSIUnit[scientific-notation=engineering,prefixes-as-symbols=false]{\ns}{\nano\second}

    \begin{document}

    \SI{10e2}{\nanosecond}
    \SI{10e2}{\ns}

    \end{document}

[LKB]

Compiled it shows different behaviour pattern for \nanosecond and \ns. It seems like I can re-define existing units and have to create new ones only. Is this correct?

[josephwright]

Settings per unit only apply to exactly that unit command. So \nanosecond and \ns are different: if you want to cahnge how \ns behaves you will need to redefine it. The reason for this is that with compound units what 'should' happen rapidly becomes impossible to tell. So siunitx only checks for a set of options for the 'outer' unit (the one you actually type in), and not for what that then is converted to.