mjhjmtu

Is it possible to automatically give names to all intersection points of a grid, so that I can draw from a specific intersection to another.

In my example, I manually gave names to the coordinates of some intersection points to illustrate my point.

So I want to draw from intersection a to intersection b, not from position (1,1) to (2,2); and from intersection x to intersection z, not from position (3,3) to (4,3)

Can this be done without using an extra package, because I can not install new packages (on my workplace computer) for security reasons.

documentclassbeamer
setbeamertemplatenavigation symbols
usepackagetikz
usepackagepgfplots
begindocument
beginframe[t]
frametitleN
begincenter
begintikzpicture
draw[step=.5cm,color=gray] (0,0) grid (8,8);
draw<2->[ultra thick, green, -latex] (1,1) -- (2,2);
draw[blue!40!black, thick, fill=blue!40!white, opacity=.5] (3,3) circle (.2cm) node (c);
draw<3->[ultra thick, red, -latex] (c.center) -- (4,3);
node at (1,1)(a)a;
node at (2,2)(b)b;
node at (3,3)(x)x;
node at (4,3)(z)z;
endtikzpicture
endcenter
endframe
enddocument

Second try at it. I hope I understood you: what I am doing here is building the grid by hand, and automatically naming the intersections. You can change the grid parameters (x and y step) and let everything else untouched; the intersections are called g-<column>-<row>.

Another possibility is to use a TikZ matrix of nodes for this.

documentclassbeamer
setbeamertemplatenavigation symbols
usepackagetikz
usepackagepgfplots
begindocument
beginframe[t]
 frametitleN
 begincenter
 begintikzpicture
 % draw[step=.5cm,color=gray] (0,0) grid (8,8);
 defmystepx0.8cm
 defmystepy0.5cm
 foreach x in 0,...,8 
 % draw rows and columns
 draw [red] (x*mystepx, 0) -- (x*mystepx,8*mystepy);
 draw [blue] (0,x*mystepy) -- (8*mystepx,x*mystepy);
 foreach y in 0,...,8 
 coordinate (g-x-y) at (x*mystepx,y*mystepy);
 
 
 %
 node at (g-1-1)a;
 node at (g-2-2)b;
 node at (g-3-3)x;
 node at (g-4-3)z;
 draw<2->[ultra thick, green, -latex] (g-1-1) -- (g-2-2);
 draw[blue!40!black, thick, fill=blue!40!white, opacity=.5] (g-3-3) circle (.2cm) node (c);
 draw<3->[ultra thick, red, -latex] (c.center) -- (g-4-3);
 endtikzpicture
 endcenter
endframe
enddocument

If you remove the units from the steps and do, say, begintikzpicture[x=2cm,y=2cm] this will increase the step of the grid while keeping the same intersections (or lattice nodes) connected.

documentclassbeamer
setbeamertemplatenavigation symbols
usepackagetikz
usepackagepgfplots
begindocument
beginframe[t]
frametitleN
begincenter
begintikzpicture[x=2cm,y=2cm]
draw[step=.5,color=gray] (0,0) grid (5,4);
draw<2->[ultra thick, green, -latex] (1,1) -- (2,2);
draw[blue!40!black, thick, fill=blue!40!white, opacity=.5] (3,3) circle (.2cm) node (c);
draw<3->[ultra thick, red, -latex] (c.center) -- (4,3);
node at (1,1)a;
node at (2,2)b;
node at (3,3)x;
node at (4,3)z;
endtikzpicture
endcenter
endframe
enddocument

Here is an example of to path style that name all intersections of a grid. When you set named grid=test the intersections will be named as test-1-1, test-1-2, ... and so on.

documentclass[tikz,border=7pt]standalone
usetikzlibrarycalc
% The x-step and the y-step lengths
newlengthdxsetlengthdx1cm
newlengthdysetlengthdy1cm
% recover the steps
makeatletter
letpgf@pathgrid@originalpgf@pathgrid
defpgf@pathgrid[#1]#2#3%
 pgfset#1%
 pgfmathsetlengthdxpgfkeysvalueof/pgf/stepx%
 globaldx=dx
 pgfmathsetlengthdypgfkeysvalueof/pgf/stepy%
 globaldy=dy
 pgf@pathgrid@original[#1]#2#3

makeatother
% define the `named grid` style
tikzset
 named grid/.style=
 to path=
 (tikztostart) grid (tikztotarget)
 let p1=(tikztostart), p2=(tikztotarget),
 n1=min(x1,x2),n2=max(x1,x2),
 n3=ceil(n1/dx),n4=n3+1,n5=int(n2/dx),
 n6=min(y1,y2),n7=max(y1,y2),
 n8=ceil(n6/dy),n9=n8+1,n10=int(n7/dy)
 in
 foreach[count=nx from 0] x in n3,n4,...,n5
 foreach[count=ny from 0] y in n8,n9,...,n10
 (x*dx,y*dy) coordinate (#1-nx-ny) % <- name the intersections
 
 
 
 ,
 named grid/.default=grid

begindocument
 begintikzpicture
 draw[rotate=35,xstep=.7,ystep=.35] (-.5,-1.5) to[named grid=test] (3.5,1.5)
 (test-1-1) edge[-latex,red,thick] (test-5-7)
 (test-0-0) circle(3pt) (test-5-8) circle(3pt);
 endtikzpicture
enddocument

Notes :

• This code is note very stable : it doesn't work well with non linear transforms.


• In a strange way, the most difficult part for me was to recover the steps.