The question is: how to use two np.where in the same statement, like this (oversimplified):
np.where((ndarr1==ndarr2),np.where((ndarr1+ndarr2==ndarr3),True,False),False)
To avoid computing second conditional statement if the first is not reached.
My first objective is to find the intersection of a ray in a triangle, if there is one. This problem can be solved by this algorithm (found on stackoverflow):
def intersect_line_triangle(q1,q2,p1,p2,p3):
def signed_tetra_volume(a,b,c,d):
return np.sign(np.dot(np.cross(b-a,c-a),d-a)/6.0)
s1 = signed_tetra_volume(q1,p1,p2,p3)
s2 = signed_tetra_volume(q2,p1,p2,p3)
if s1 != s2:
s3 = signed_tetra_volume(q1,q2,p1,p2)
s4 = signed_tetra_volume(q1,q2,p2,p3)
s5 = signed_tetra_volume(q1,q2,p3,p1)
if s3 == s4 and s4 == s5:
n = np.cross(p2-p1,p3-p1)
t = np.dot(p1-q1,n) / np.dot(q2-q1,n)
return q1 + t * (q2-q1)
return None
Here are two conditional statements:
s1!=s2
s3==s4 & s4==s5
Now since I have >20k triangles to check, I want to apply this function on all triangles at the same time.
First solution is:
s1 = vol(r0,tri[:,0,:],tri[:,1,:],tri[:,2,:])
s2 = vol(r1,tri[:,0,:],tri[:,1,:],tri[:,2,:])
s3 = vol(r1,r2,tri[:,0,:],tri[:,1,:])
s4 = vol(r1,r2,tri[:,1,:],tri[:,2,:])
s5 = vol(r1,r2,tri[:,2,:],tri[:,0,:])
np.where((s1!=s2) & (s3+s4==s4+s5),intersect(),False)
where s1,s2,s3,s4,s5 are arrays containing the value S for each triangle. Problem is, it means I have to compute s3,s4,and s5 for all triangles.
Now the ideal would be to compute statement 2 (and s3,s4,s5) only when statement 1 is True, with something like this:
check= np.where((s1!=s2),np.where((compute(s3)==compute(s4)) & (compute(s4)==compute(s5), compute(intersection),False),False)
(to simplify explanation, I just stated 'compute' instead of the whole computing process. Here, 'compute' is does only on the appropriate triangles).
Now of course this option doesn't work (and computes s4 two times), but I'd gladly have some recommendations on a similar process
Here's how I used masked arrays to answer this problem:
loTrue= np.where((s1!=s2),False,True)
s3=ma.masked_array(np.sign(dot(np.cross(r0r1, r0t0), r0t1)),mask=loTrue)
s4=ma.masked_array(np.sign(dot(np.cross(r0r1, r0t1), r0t2)),mask=loTrue)
s5=ma.masked_array(np.sign(dot(np.cross(r0r1, r0t2), r0t0)),mask=loTrue)
loTrue= ma.masked_array(np.where((abs(s3-s4)<1e-4) & ( abs(s5-s4)<1e-4),True,False),mask=loTrue)
#also works when computing s3,s4 and s5 inside loTrue, like this:
loTrue= np.where((s1!=s2),False,True)
loTrue= ma.masked_array(np.where(
(abs(np.sign(dot(np.cross(r0r1, r0t0), r0t1))-np.sign(dot(np.cross(r0r1, r0t1), r0t2)))<1e-4) &
(abs(np.sign(dot(np.cross(r0r1, r0t2), r0t0))-np.sign(dot(np.cross(r0r1, r0t1), r0t2)))<1e-4),True,False)
,mask=loTrue)
Note that the same process, when not using such approach, is done like this:
s3= np.sign(dot(np.cross(r0r1, r0t0), r0t1) /6.0)
s4= np.sign(dot(np.cross(r0r1, r0t1), r0t2) /6.0)
s5= np.sign(dot(np.cross(r0r1, r0t2), r0t0) /6.0)
loTrue= np.where((s1!=s2) & (abs(s3-s4)<1e-4) & ( abs(s5-s4)<1e-4) ,True,False)
Both give the same results, however, when looping on this process only for 10k iterations, NOT using masked arrays is faster! (26 secs without masked arrays, 31 secs with masked arrays, 33 when using masked arrays in one line only (not computing s3,s4 and s5 separately, or computing s4 before).
Conclusion: using nested arrays is solved here (note that the mask indicates where it won't be computed, hence first loTri must bet set to False (0) when condition is verified). However, in that scenario, it's not faster.
I can get a small speedup from short circuiting but I'm not convinced it is worth the additional admin.
full computation 4.463818839867599 ms per iteration (one ray, 20,000 triangles)
short ciruciting 3.0060838296776637 ms per iteration (one ray, 20,000 triangles)
Code:
import numpy as np
def ilt_cut(q1,q2,p1,p2,p3):
qm = (q1+q2)/2
qd = qm-q2
p12 = p1-p2
aux = np.cross(qd,q2-p2)
s3 = np.einsum("ij,ij->i",aux,p12)
s4 = np.einsum("ij,ij->i",aux,p2-p3)
ge = (s3>=0)&(s4>=0)
le = (s3<=0)&(s4<=0)
keep = np.flatnonzero(ge|le)
aux = p1[keep]
qpm1 = qm-aux
p31 = p3[keep]-aux
s5 = np.einsum("ij,ij->i",np.cross(qpm1,p31),qd)
ge = ge[keep]&(s5>=0)
le = le[keep]&(s5<=0)
flt = np.flatnonzero(ge|le)
keep = keep[flt]
n = np.cross(p31[flt], p12[keep])
s12 = np.einsum("ij,ij->i",n,qpm1[flt])
flt = np.abs(s12) <= np.abs(s3[keep]+s4[keep]+s5[flt])
return keep[flt],qm-(s12[flt]/np.einsum("ij,ij->i",qd,n[flt]))[:,None]*qd
def ilt_full(q1,q2,p1,p2,p3):
qm = (q1+q2)/2
qd = qm-q2
p12 = p1-p2
qpm1 = qm-p1
p31 = p3-p1
aux = np.cross(qd,q2-p2)
s3 = np.einsum("ij,ij->i",aux,p12)
s4 = np.einsum("ij,ij->i",aux,p2-p3)
s5 = np.einsum("ij,ij->i",np.cross(qpm1,p31),qd)
n = np.cross(p31, p12)
s12 = np.einsum("ij,ij->i",n,qpm1)
ge = (s3>=0)&(s4>=0)&(s5>=0)
le = (s3<=0)&(s4<=0)&(s5<=0)
keep = np.flatnonzero((np.abs(s12) <= np.abs(s3+s4+s5)) & (ge|le))
return keep,qm-(s12[keep]/np.einsum("ij,ij->i",qd,n[keep]))[:,None]*qd
tri = np.random.uniform(1, 10, (20_000, 3, 3))
p0, p1 = np.random.uniform(1, 10, (2, 3))
from timeit import timeit
A,B,C = tri.transpose(1,0,2)
print('full computation', timeit(lambda: ilt_full(p0[None], p1[None], A, B, C), number=100)*10, 'ms per iteration (one ray, 20,000 triangles)')
print('short ciruciting', timeit(lambda: ilt_cut(p0[None], p1[None], A, B, C), number=100)*10, 'ms per iteration (one ray, 20,000 triangles)')
Note that I played a bit with the algorithm, so this may not in every edge case give the same result aas yours.
What I changed:
I inlined the tetra volume, which allows to save a few repeated subcomputations
I replace one of the ray ends with the midpoint M of the ray. This saves computing one tetra volume (s1 or s2) because one can check whether the ray crosses the triangle ABC plane by comparing the volume of tetra ABCM to the sum of s3, s4, s5 (if they have the same signs).
I am doing problem 61 at project Euler and came up with the following code (to test the case they give):
p3 n = n*(n+1) `div` 2
p4 n = n*n
p5 n = n*(3*n -1) `div` 2
p6 n = n*(2*n -1)
p7 n = n*(5*n -3) `div` 2
p8 n = n*(3*n -2)
x n = take 2 $ show n
x2 n = reverse $ take 2 $ reverse $ show n
pX p = dropWhile (< 999) $ takeWhile (< 10000) [p n|n<-[1..]]
isCyclic2 (a,b,c) = x2 b == x c && x2 c == x a && x2 a == x b
ns2 = [(a,b,c)|a <- pX p3 , b <- pX p4 , c <- pX p5 , isCyclic2 (a,b,c)]
And all ns2 does is return an empty list, yet cyclic2 with the arguments given as the example in the question, yet the series doesn't come up in the solution. The problem must lie in the list comprehension ns2 but I can't see where, what have I done wrong?
Also, how can I make it so that the pX only gets the pX (n) up to the pX used in the previous pX?
PS: in case you thought I completely missed the problem, I will get my final solution with this:
isCyclic (a,b,c,d,e,f) = x2 a == x b && x2 b == x c && x2 c == x d && x2 d == x e && x2 e == x f && x2 f == x a
ns = [[a,b,c,d,e,f]|a <- pX p3 , b <- pX p4 , c <- pX p5 , d <- pX p6 , e <- pX p7 , f <- pX p8 ,isCyclic (a,b,c,d,e,f)]
answer = sum $ head ns
The order is important. The cyclic numbers in the question are 8128, 2882, 8281, and these are not P3/127, P4/91, P5/44 but P3/127, P5/44, P4/91.
Your code is only checking in the order 8128, 8281, 2882, which is not cyclic.
You would get the result if you check for
isCyclic2 (a,c,b)
in your list comprehension.
EDIT: Wrong Problem!
I assumed you were talking about the circular number problem, Sorry!
There is a more efficient way to do this with something like this:
take (2 * l x -1) . cycle $ show x
where l = length . show
Try that and see where it gets you.
If I understand you right here, you're no longer asking why your code doesn't work but how to make it faster. That's actually the whole fun of Project Euler to find an efficient way to solve the problems, so proceed with care and first try to think of reducing your search space yourself. I suggest you let Haskell print out the three lists pX p3, pX p4, pX p5 and see how you'd go about looking for a cycle.
If you would proceed like your list comprehension, you'd start with the first element of each list, 1035, 1024, 1080. I'm pretty sure you would stop right after picking 1035 and 1024 and not test for cycles with any value from P5, let alone try all the permutations of the combinations involving these two numbers.
(I haven't actually worked on this problem yet, so this is how I would go about speeding it up. There may be some math wizardry out there that's even faster)
First, start looking at the numbers you get from pX. You can drop more than those. For example, P3 contains 6105 - there's no way you're going to find a number in the other sets starting with '05'. So you can also drop those numbers where the number modulo 100 is less than 10.
Then (for the case of 3 sets), we can sometimes see after drawing two numbers that there can't be any number in the last set that will give you a cycle, no matter how you permutate (e.g. 1035 from P3 and 3136 from P4 - there can't be a cycle here).
I'd probably try to build a chain by starting with the elements from one list, one by one, and for each element, find the elements from the remaining lists that are valid successors. For those that you've found, continue trying to find the next chain element from the remaining lists. When you've built a chain with one number from every list, you just have to check if the last two digits of the last number match the first two digits of the first number.
Note when looking for successors, you again don't have to traverse the entire lists. If you're looking for a successor to 3015 from P5, for example, you can stop when you hit a number that's 1600 or larger.
If that's too slow still, you could transform the lists other than the first one to maps where the map key is the first two digits and the associated values are lists of numbers that start with those digits. Saves you from going through the lists from the start again and again.
I hope this helps a bit.
btw, I sense some repetition in your code.
you can unite your [p3, p4, p5, p6, p7, p8] functions into one function that will take the 3 from the p3 as a parameter etc.
to find what the pattern is, you can make all the functions in the form of
pX n = ... `div` 2