5
$\begingroup$

I would like to draw a Calabi-Yau space in Mathematica. From Andrew J. Hanson's paper (page 6) about this topic, I got the following code, which just does not show anything at all in my Mathematica test version. I am a newbie coming from the processing/open frameworks direction. Any help would be highly appreciated - Thank you!

cCos[theta_, xi_] := .5 (Eˆ (xi + I theta) + Eˆ (-xi - I theta));

cSin[theta_, xi_] := (-.5 I) (Eˆ (xi + I theta) - Eˆ (-xi - I theta));

z1[theta_, xi_, n_, k_] := Eˆ (k*2*Pi*I/n)*cCos[theta, xi] ˆ (2.0/n);

z2[theta_, xi_, n_, k_] := Eˆ (k*2*Pi*I/n)*cSin[theta, xi] ˆ (2.0/n);

pz1[theta_, xi_, n_, k_] := Eˆ ((xi + I theta)/n)*Eˆ (k*2*Pi*I/n);

pz2[theta_, xi_, n_, k_] := Eˆ ((-xi - I theta)/n)*Eˆ (-k*2*Pi*I/n);

MakePolygons[vl_List] := 
 Block[{l = vl, l1 = Map[RotateLeft, vl], mesh}, 
  mesh = {l, l1, RotateLeft[l1], RotateLeft[l]};
  mesh = Map[Drop[#, -1] &, mesh, {1}];
  mesh = Map[Drop[#, -1] &, mesh, {2}];
  Polygon /@ Transpose[Map[Flatten[#, 1] &, mesh]]]

n1 = 3; n2 = 3; xiSteps = 17; xiMax = 1; thetaSteps = 17; angle = Pi/4;

cosA = Cos[angle]; sinA = Sin[angle];

Do[Do[patch33[k1 + 1, k2 + 1] = 
    MakePolygons[
     Table[Block[{z1Val = N[z1[theta, xi, n1, k1]], 
        z2Val = N[z2[theta, xi, n2, k2]]}, {Re[z1Val], Re[z2Val], 
        cosA*Im[z1Val] + sinA*Im[z2Val]}], {xi, -xiMax, 
       xiMax, (2*xiMax)/(xiSteps - 1)}, {theta, 0, 
       Pi/2, (Pi/2)/(thetaSteps - 1)}]], {k1, 0, n1 - 1}], {k2, 0, 
   n2 - 1}];
zi -> pzi and Pi/2 -> 2 Pi;

bs0 = 0.8; bs1 = 0.2; lt = 0.9;

surface33 = 
  Show[Graphics3D[
    Table[[Block[{bs = 
          If[And[k1 == 0, k2 == 0], bs0, bs1]}, {RGBColor[
          bs + lt*k1/(n1 - 1), bs + lt*k2/(n2 - 1), bs], 
         patch33[k1 + 1, k2 + 1]}] {k1, 0, n1 - 1}, {k2, 0, n2 - 1}], 
     Lighting -> False, Axes -> None, Boxed -> False, 
     BoxRatios -> {1, 1, 1}, ViewPoint -> {2.9, 1.0, 1.4}]]];
$\endgroup$

2 Answers 2

13
$\begingroup$

There are many mistakes which I believe come from the Copy/Paste:

cCos[theta_, xi_] := .5 (E^(xi + I theta) + E^ (-xi - I theta));
cSin[theta_, xi_] := (-.5 I) (E^ (xi + I theta) - E^ (-xi - I theta));
z1[theta_, xi_, n_, k_] := E^ (k*2*Pi*I/n)*cCos[theta, xi] ^ (2.0/n);
z2[theta_, xi_, n_, k_] := E^ (k*2*Pi*I/n)*cSin[theta, xi] ^ (2.0/n);
pz1[theta_, xi_, n_, k_] := E^ ((xi + I theta)/n)*E^ (k*2*Pi*I/n);
pz2[theta_, xi_, n_, k_] := E^ ((-xi - I theta)/n)*E^ (-k*2*Pi*I/n);
MakePolygons[vl_List] := 
 Block[{l = vl, l1 = Map[RotateLeft, vl], mesh}, 
  mesh = {l, l1, RotateLeft[l1], RotateLeft[l]};
  mesh = Map[Drop[#, -1] &, mesh, {1}];
  mesh = Map[Drop[#, -1] &, mesh, {2}];
  Polygon /@ Transpose[Map[Flatten[#, 1] &, mesh]]]

n1 = 3; n2 = 3; xiSteps = 17; xiMax = 1; thetaSteps = 17; angle = Pi/4;

cosA = Cos[angle]; sinA = Sin[angle];

Do[Do[patch33[k1 + 1, k2 + 1] = 
    MakePolygons[
     Table[Block[{z1Val = N[z1[theta, xi, n1, k1]], 
        z2Val = N[z2[theta, xi, n2, k2]]}, {Re[z1Val], Re[z2Val], 
        cosA*Im[z1Val] + sinA*Im[z2Val]}], {xi, -xiMax, 
       xiMax, (2*xiMax)/(xiSteps - 1)}, {theta, 0, 
       Pi/2, (Pi/2)/(thetaSteps - 1)}]], {k1, 0, n1 - 1}], {k2, 0, n2 - 1}];

bs0 = 0.8; bs1 = 0.2; lt = 0.9;

surface33 = 
 Show[Graphics3D[
   Table[Block[{bs = If[And[k1 == 0, k2 == 0], bs0, bs1]}, {RGBColor[
       bs + lt*k1/(n1 - 1), bs + lt*k2/(n2 - 1), bs], 
      patch33[k1 + 1, k2 + 1]}], {k1, 0, n1 - 1}, {k2, 0, n2 - 1}], 
   Lighting -> Automatic, Axes -> None, Boxed -> False, 
   BoxRatios -> {1, 1, 1}, ViewPoint -> {2.9, 1.0, 1.4}]]

Mathematica graphics

$\endgroup$
1
  • 1
    $\begingroup$ Dear Öska - awesome! thank you very much for fixing it so fast. THX!!!! $\endgroup$
    – MarkusKla
    Oct 8, 2014 at 14:39
10
$\begingroup$

Considering the use of the old utility MakePolygons[] by Roman Maeder, as well as the year Hanson's paper appeared, I believe this was done during the time one still had to load a package to be able to use ParametricPlot3D[]. Since ParametricPlot3D[] has been built-in for quite a while now, please allow me to present a modernized plot of the Fermat surface for $x^3 + y^3 = 1$:

z1[θ_, ξ_, n_, k_] := Exp[2 π I k/n] Cosh[ξ + I θ]^(2/n);
z2[θ_, ξ_, n_, k_] := Exp[2 π I k/n] Sinh[ξ + I θ]^(2/n);

With[{n1 = 3, n2 = 3, φ = π/4, bs0 = 0.8, bs1 = 0.2, lt = 0.9}, 
     ParametricPlot3D[Flatten[Table[
                      With[{z1Val = z1[θ, ξ, n1, k1], z2Val = z2[θ, ξ, n2, k2]},
                           {Re[z1Val], Re[z2Val],
                            Cos[φ] Im[z1Val] + Sin[φ] Im[z2Val]}],
                                    {k1, 0, n1 - 1}, {k2, 0, n2 - 1}], 1],
                      {ξ, -1, 1}, {θ, 0, π/2}, Axes -> None, Boxed -> False,
                      Evaluated -> True, Lighting -> "Neutral", 
                      PlotStyle -> Flatten[Table[
                      With[{b = bs1 + (bs0 - bs1) Boole[k1 == k2 == 0]}, 
                           RGBColor[b + lt k1/(n1 - 1), b + lt k2/(n2 - 1), b]],
                          {k1, 0, n1 - 1}, {k2, 0, n2 - 1}]],
                      ViewPoint -> {2.9, 1.0, 1.4}]]

Calabi-Yau manifold for x^3 + y^3 = 1

Here's a plot of a different Calabi-Yau manifold; I'll let you guess what parameters I used to generate this: guess this picture

$\endgroup$
3
  • $\begingroup$ M., is there a copyright in your image? i.stack.imgur.com/3csGv.png Can we use it? Thank you $\endgroup$
    – francis
    Mar 25, 2019 at 12:34
  • $\begingroup$ @francis, you are free to use it; all that's needed is for you to link to this post for credit. $\endgroup$ Mar 25, 2019 at 14:20
  • $\begingroup$ @JM., thank you for your reply $\endgroup$
    – Francis
    Mar 25, 2019 at 17:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.