Hello!
I have the following program:
from visual import *
def frange(start, end=None, inc=None): #"A range function, that does accept float increments..."
if end == None:
end = start + 0.0
start = 0.0
if inc == None:
inc = 1.0
L = []
while 1:
next = start + len(L) * inc
if inc > 0 and next >= end:
break
elif inc < 0 and next <= end:
break
L.append(next)
return L
scene2 = display(title='Naboja',
x=0, y=0, width=1380, height=950,
center=(0,0,0), background=(0,0,0))
def obroc_tocke_narisi(): #Draw points of the charged ring
global tocke
tocke = []
q = 0.004
for y in frange(-1.1,1.1,0.01):
for z in frange (-1.1,1.1, 0.01):
if y**2 + z**2 <= (1+q)**2 and y**2 + z**2 >= (1-q)**2 :
tocke.append([0, y, z])
points(pos=(0,y,z), color=color.red, size=0.000001)
global n_obroc
n_obroc = len(tocke)
print "Koncan seznam tock obroca"
obroc_tocke_narisi()
def puscice(): #Show the vector field of the points
for a in frange(-4,4,0.15):
for b in frange (-2.6, 2.6, 0.15):
for z in frange (1):
v = vector(0,0,0)
for c in range(n_obroc):
v += (vector(a,b,z) - vector(tocke[c]))/(mag(vector(a,b,z) - vector(tocke[c])))**2
arrow(axis=0.0015*v, pos=(a, b,z), shaftwidth=0.01, color=color.white)
print "Koncan vektorski prikaz polja"
puscice()
class Tocka: #Calculate a point cloud. Make an array of Tocka objects which have coordinates and potential
def __init__(self, koordinate, potencial):
self.koordinate = koordinate
self.potencial = potencial
def oblak():
global t
t = []
for a in frange(-4,4,0.01):
for b in frange (-2.6, 2.6, 0.01):
for z in frange (-0.7,0,0.01):
t.append(Tocka([a,b,z], 0))
print "Koncano racunanje oblaka tock"
oblak()
global ravnina
ravnina = len(t)
def potenciali(): #Calculate potentials for every point
for s in range (ravnina):
for l in range (n_obroc):
z = mag(vector(t[s].koordinate) - vector(tocke[l]))
if z != 0:
t[s].potencial = t[s].potencial + 0.01/z
print "Koncano racunanje potencialov"
potenciali()
def ekvipotencialke_0(): #Draw points of same potential and color them
for q in frange (0, 0.1, 0.1):
for s in range(ravnina):
if q >= round(t[s].potencial,3)-0.01 and q <= round(t[s].potencial,3) + 0.01 :
points(pos=(t[s].koordinate), size=0.1, color = color.blue)
def ekvipotencialke_1():
for q in frange (0, 1, 0.5):
for s in range(ravnina):
if q >= round(t[s].potencial,3)-0.01 and q <= round(t[s].potencial,3) + 0.01 :
points(pos=(t[s].koordinate), size=0.1, color = color.blue)
def ekvipotencialke_2():
for q in frange (1, 2, 0.5):
for s in range(ravnina):
if q >= round(t[s].potencial,3)-0.01 and q <= round(t[s].potencial,3) + 0.01 :
points(pos=(t[s].koordinate), size=0.1, color = color.blue)
def ekvipotencialke_3():
for q in frange (2, 3, 0.5):
for s in range(ravnina):
if q >= round(t[s].potencial,3)-0.01 and q <= round(t[s].potencial,3) + 0.01 :
points(pos=(t[s].koordinate), size=0.1, color = color.orange)
def ekvipotencialke_4():
for q in frange (3,4, 0.5):
for s in range(ravnina):
if q >= round(t[s].potencial,3)-0.01 and q <= round(t[s].potencial,3) + 0.01 :
points(pos=(t[s].koordinate), size=0.1, color = color.cyan)
def ekvipotencialke_5():
for q in frange (4,5, 0.5):
for s in range(ravnina):
if q >= round(t[s].potencial,3)-0.01 and q <= round(t[s].potencial,3) + 0.01 :
points(pos=(t[s].koordinate), size=0.1, color = color.magenta)
def ekvipotencialke_6():
for q in frange (5,6, 0.5):
for s in range(ravnina):
if q >= round(t[s].potencial,3)-0.01 and q <= round(t[s].potencial,3) + 0.01 :
points(pos=(t[s].koordinate), size=0.1, color = color.yellow)
ekvipotencialke_0()
ekvipotencialke_1()
ekvipotencialke_2()
ekvipotencialke_3()
ekvipotencialke_4()
ekvipotencialke_5()
ekvipotencialke_6()
print "Koncano risanje ekvipotencialk"
So, if you set z to zero in the oblak() function and leave it overnight, it would make a beautiful calculation of equipotential points near a charged ring thet looks something like the picture I've included. (You need the Visual python library).
So, now I want to go 3D (this would spread the circles into spheres, should be really beautiful). If I run this, I will quickly run out of memory. So, firstly:
-How do I get a PP job to return an array and make it global. I.E, how would I split the oblak() function, submit it to PP and get two returned global arrays, wait for the job to finish, then merge the arrays. This applies to variables as well. How do you work with and return global variables in PP?-I kindly request you, Vitali, to look (just look) at the problem of making PP work with visual in such a way that many PP processes draw in one Visual window at once. Currently, this isn't possible. I've contacted Bruce Sherwood, the author of Visual and he says that I can't probably achieve this. If i try multiprocessing, it keeps spawning new windows until it crashes. If I use threading, it still only runs one thread (because of the global interpreter lock), although it still is faster. Using PP, nothing happens. Here's the code:
from visual import *
from threading import *
def frange(start, end=None, inc=None):
"A range function, that does accept float increments..."
if end == None:
end = start + 0.0
start = 0.0
if inc == None:
inc = 1.0
L = []
while 1:
next = start + len(L) * inc
if inc > 0 and next >= end:
break
elif inc < 0 and next <= end:
break
L.append(next)
return L
scene2 = display(title='Naboja',
x=0, y=0, width=1000, height=800,
center=(0,0,0), background=(0,0,0))
naboj1 = sphere(pos=(-1,0,0), radius = 0.05, color = color.red)
naboj2 = sphere(pos=(1,0,0), radius = 0.05 , color = color.blue)
naboj1.naboj = 0.165
naboj2.naboj = -1
for a in frange(-2, 2, 0.1):
for b in frange (-1, 1, 0.1):
ri = vector(a,b,0)
r1 = naboj1.pos - ri
r2 = naboj2.pos - ri
v1 = r1 * naboj1.naboj / (mag(r1)**2)
v2 = r2 * naboj2.naboj / (mag(r2)**2)
v = v1 + v2
arrow (axis=0.1*norm(v), pos=(a,b,0) , color = color.red, shaftwidth=0.0000001)
def potenciali1():
scene2.select()
for q in frange (0, 5, 0.5):
for a in frange(-2, 2, 0.01):
for b in frange (-1, 1, 0.01):
ri = vector(a,b,0)
r1 = naboj1.pos - ri
r2 = naboj2.pos - ri
if mag(r1) != 0 and mag(r2) != 0:
p1 = naboj1.naboj/mag(r1)
p2 = naboj2.naboj/mag(r2)
p = p1 + p2
if q >= round(p,3)-0.01 and q <= round(p,1)+0.01:
points(pos=(a,b,0), size=0.01, color = color.green)
def potenciali2():
scene2.select()
for w in frange (-5, 0, 0.5):
for c in frange(-2, 2, 0.01):
for d in frange (-1, 1, 0.01):
ri1 = vector(c,d,0)
r11 = naboj1.pos - ri1
r21 = naboj2.pos - ri1
if mag(r11) != 0 and mag(r21) != 0:
p11 = naboj1.naboj/mag(r11)
p21 = naboj2.naboj/mag(r21)
s = p11 + p21
if w >= round(s,3)-0.01 and w <= round(s,1)+0.01:
points(pos=(c,d,0), size=0.01, color = color.yellow)
delo1 = Thread(target=potenciali1)
delo2 = Thread(target=potenciali2)
delo1.start()
delo2.start()
So, again, how could I get this to work with PP and Visual. Vitalii, since you are experienced in this, could you maybe even contact Bruce (getting Visual to run multiple drawing threads in the same window at once - the effect I achieve here with the yellow and green points) Thank you :-D
-Lastly, for my first problem, does anybody have a rough idea on how many computers/memory I would need to sucessfully pull the 3D version off? That is according to my calculations 29.120.000 points, each with 4 sets of floating point data (coordinates x,y,z and scalar potential)
Forums 
