**A** geometric Le Monde mathematical puzzle:

*Given a pizza of diameter 20cm, what is the way to cut it by two perpendicular lines through a point distant 5cm from the centre towards maximising the surface of two opposite slices? *
* Using the same point as the tip of the four slices, what is the way to make four slices with equal arcs in four cuts from the tip again towards maximising the surface of two opposite slices?*

**F**or both questions, I did not bother with the maths but went itself to a discretisation of the disk, counting the proportion of points within two opposite slices and letting the inclination of these slices move from zero to π/2. Unsurprisingly, for the first question, the answer is π/4, given that there is no difference between both surfaces at angles 0 and π/2. My R code is as follows, using (5,0) as the tip:

M=100
surfaz=function(alpha){
surfz=0
cosal=cos(alpha);sinal=sin(alpha)
X=Y=seq(-10,10,le=M)
Xcosal=(X-5)*cosal
Xsinal=(X-5)*sinal
for (i in 1:M){
norm=sqrt(X[i]^2+Y^2)
scal1=Xsinal[i]+Y*cosal
scal2=-Xcosal[i]+Y*sinal
surfz=surfz+sum((norm<=10)*(scal1*scal2>0))}
return(4*surfz/M/M/pi)}

The second puzzle can be solved by a similar code, except that the slice area between two lines has to be determined by a cross product:

surfoz=function(alpha,ploz=FALSE){
sinal=sin(alpha);cosal=cos(alpha)
X=Y=seq(-10,10,le=M)
frsterm=cosal*(10*cosal-5)+sinal*(10*sinal-5)
trdterm=cosal*(10*cosal+5)+sinal*(10*sinal+5)
surfz=0
for (i in 1:M){
norm=sqrt(X[i]^2+Y^2)
scal1=(10*(Y[i]-5)*cosal-(10*sinal-5)*X)*frsterm
scal2=-(-10*(Y[i]-5)*sinal-(10*cosal-5)*X)*frsterm
scal3=(-10*(Y[i]-5)*cosal+(10*sinal+5)*X)*trdterm
scal4=-(10*(Y[i]-5)*sinal+(10*cosal+5)*X)*trdterm
surfz=surfz+sum((norm<=10)*
((scal1>0)*(scal2>0)+
(scal3>0)*(scal4>0)))}
return(4*surfz/M/M/pi)}

a code that shows that *all* cuts lead to identical surfaces for bot sets of slices. A fairly surprising result!