diff options
author | Norbert Preining <norbert@preining.info> | 2023-08-23 03:01:02 +0000 |
---|---|---|
committer | Norbert Preining <norbert@preining.info> | 2023-08-23 03:01:02 +0000 |
commit | 76d75859475de5b0b64017c9dfc11644a2ece537 (patch) | |
tree | e2f334df5d811d8fe54215e0591eaf227de58fe0 /fonts/notomath/doc/notomath-doc.tex | |
parent | cf3e954514eb50bea0335bdc97d6b76f80f5d03e (diff) |
CTAN sync 202308230300
Diffstat (limited to 'fonts/notomath/doc/notomath-doc.tex')
-rw-r--r-- | fonts/notomath/doc/notomath-doc.tex | 27 |
1 files changed, 20 insertions, 7 deletions
diff --git a/fonts/notomath/doc/notomath-doc.tex b/fonts/notomath/doc/notomath-doc.tex index f84df86388..e985a7ab30 100644 --- a/fonts/notomath/doc/notomath-doc.tex +++ b/fonts/notomath/doc/notomath-doc.tex @@ -8,7 +8,7 @@ %\usepackage{amssymb}% don't use with newtxmath %SetFonts % libertine+newtxmath -\pdfmapfile{=NotoMath.map} +%\pdfmapfile{=NotoMath.map} \usepackage[T1]{fontenc} %\usepackage[scaled=1.1]{zlmtt} %\usepackage[scaled=1.1]{newtxtt} @@ -55,7 +55,7 @@ The effect of this is: load {\tt noto-serif} and {\tt noto-sans} scaled down by the factor .9 to an x-height of 482; \item set the main text font to NotoSerif and set \verb|\sfdefault| to NotoSans; -\item the only weight used from the nine available weights are {\tt regular} and {\tt bold}, as these are the weights used in {\tt newtxmath} with options {\tt noto} and {\tt notosans}; +\item the only weights used from the nine available weights are {\tt regular} and {\tt bold}, as these are the weights used in {\tt newtxmath} with options {\tt noto} and {\tt notosans}; \item load {\tt newtxmath} with option {noto} at natural scale. \end{itemize} @@ -90,7 +90,8 @@ and works well in {\tt T1} encoding. There is no {\tt OT1} encoded version curre \item The TT package {\tt nimbusmononarrow} does have a {\tt visiblespace} glyph and its {\tt OT1} encoding is in {\tt TEX TYPEWRITER TEXT} so \verb|\verb| and \verb|\texttt| function correctly even in {\tt OT1} encoding. Caps are not too tall---this is my preferred serifed TT with NotoSans text. \end{itemize} -\item +\item As with all {\tt newtxmath} based packages, you should not load {\tt amssymb}---{\tt newtxmath} has its own replacement for the AMSm math symbol fonts, of a more approriate size and color match. You do not need to load {\tt amsmath}, though doing so does no harm. + \end{itemize} @@ -126,12 +127,24 @@ will output the Noto fonts scaled to x-height 482 with matching math symbols. Th \textsc{Example 4:} \begin{verbatim} \usepackage[scaled=1.24]{nimbusmononarrow}% typewriter font -\usepackage[scale=1.11,sfdefault,pf,osf]{notomath} +\usepackage[scale=1.11,sfdefault,pf,osf,subscriptcorrection]{notomath} % load NotoSerif, NotoSans, mainfont=NotoSans % option subscriptcorrection passed to newtxmath \end{verbatim} will output the Noto fonts scaled to x-height $536$ with matching math symbols. The main text font will be NotoSans with proportional oldstyle figures except in math, which always uses tabular lining figures. +\textsc{Example 5:} +\begin{verbatim} +\usepackage[scaled=1.24]{nimbusmononarrow}% typewriter font +\usepackage[scale=1.11,sfdefault,pf,osf,subscriptcorrection]{notomath} +% load NotoSans as mainfont, then newtxmath with option notosans +% option subscriptcorrection passed to newtxmath +\usepackage[book,medium,scaled=1,sfdefault]{FiraSans} +% then replace mainfont by FiraSans +\end{verbatim} +will output FiraSans as main text font, with math symbols from {\tt newtxmath} and math letters from {\tt notosans}. This combination seems to work well with {\tt beamer}, though each {\tt scale} should be multiplied by the same factor. + + The examples above all work with {\tt pdflatex}, and with {\tt xelatex} if some additional rules are followed. With {\tt xelatex}, the lines in the above examples must precede the loading of {\tt fontspec}, which must use the option {\tt nomath}. After that, one may load any text fonts required for secondary use, or even replace the main Noto fonts. \section*{Subscript Correction} The spacing of math letters was adjusted so the superscripts would not collide with the base letters. This was necessary mainly for letters like $j$, $f$, $y$ and $\beta$ as superscripts and like $D$ and $\Omega$ as base letters. As a result of these adjustments, some of the formerly problematic superscript letters become problematic subscript letters. Two files are provided to make adjustments to the letter by inserting appropriate kerns when that letter is the first character in a subscript---one for NotoSerif and one for NotoSans letters, under the respective names @@ -209,7 +222,7 @@ At its lowest level, you invoke NotoMath in {\tt newtxmath} using the option {\t \def\e{\varepsilon} \textbf{An inversion formula:} Let $g:\mathbb{R}^+\to \mathbb{R}$ be bounded and right continuous, and let -$\varphi(\alpha)\coloneqq\int_0^\infty e^{-\alpha t}g(t)\,dt$ denote its Laplace transform. Then, for every $t>0$, +$\varphi(\alpha)\coloneq\int_0^\infty e^{-\alpha t}g(t)\,dt$ denote its Laplace transform. Then, for every $t>0$, \begin{equation} g(t)=\lim_{\mathstrut\e\to 0}\lim_{\mathstrut\lambda\to\infty}\e^{-1}\sum_{\lambda t<k\le (\lambda+\e)t} \frac{(-1)^k}{k!}\lambda^k\varphi ^{(k)}(\lambda). @@ -221,7 +234,7 @@ g(t)=\lim_{\mathstrut\e\to 0}\lim_{\mathstrut\lambda\to\infty}\e^{-1}\sum_{\lamb \end{equation} Fix an initial point $\bp_0$ in the interior of $U$, and assume $\bv(\bp_0,\bal_0)\neq\mathbf{0}$. Then, for sufficiently small $t$, $|\bp-\bp_0|$ and $|\bal-\bal_0|$, the system~(\ref{eq:first}) has a unique solution $\bx_{\bal}(t)$ satisfying the initial condition $\bx_{\bal}(0)=\bp$, and that solution depends differentiably (of class~$C^k$) on $t$, $\bp$ and $\bal$. -\textbf{Stirling's formula:} +\textbf{Stirling's formula:} \begin{equation} \Gamma(z)\sim e^{-z}z^{z-1/2}\sqrt{2\pi}\bigg[1+\frac{1}{12z}+\frac{1}{288z^2}-\frac{139}{51840z^3}+\ldots\bigg],\quad z\to\infty \text{ in $|\arg z|<\pi$}. \end{equation} @@ -231,7 +244,7 @@ $z_2$, $z_3$, $z_4$ in $\mathbb{C}$, define the B\'ezier curve with control points $z_1$, $z_2$, $z_3$, $z_4$ by -\[z(t)\coloneqq(1-t)^3z_1+3(1-t)^2tz_2+3(1-t)t^2z_3+t^3z_4,\qquad 0\le t\le 1.\] +\[z(t)\coloneq(1-t)^3z_1+3(1-t)^2tz_2+3(1-t)t^2z_3+t^3z_4,\qquad 0\le t\le 1.\] Because $(1-t)^3+3(1-t)^2t+3(1-t)t^2+t^3=(1-t+t)^3=1$ and all summands are positive for $0\le t\le1$, $z(t)$ is a convex combination of the four points~$z_k$, hence |