Wrong cosine values in Python

Question

I'm doing a simulation of an integral by taking numbers between 0 and 1, so I can get a force components. But I always get wrong cosine values, for example, when I take a=1 (so it makes a 45º angle, and therefore "L" should be equal to "k"), I get a "k" bigger than "L".If someone can please look what's wrong I'd really apreciate.

from math import pi, sin,cos, atan
for n in s:
    f=atan(n/a)
    L=ko*(Q/N*)*cos(f)
    k=ko*(Q/N)*sin(f)
    y.append(k)
    x.append(L)
yy=sum(y)
xx=sum(x)
print (xx,yy)

http://ericlippert.com/2014/03/05/how-to-debug-small-programs/ — Konstantin, Jun 16 '15 at 16:10
I am getting same values for `L` and `k`, can you share the actual program you are using with values of the variables. — DarthSpeedious, Jun 16 '15 at 16:21
Can you please edit this in the question. Not for me, but other people might find it easier to read your code. — DarthSpeedious, Jun 16 '15 at 16:27

score 5 · Answer 1 · answered Jun 16 '15 at 16:11

5

You're aware of the fact that it's in radians, right?

In [4]: math.cos(45)
Out[4]: 0.5253219888177297

In [5]: math.cos(45 / 180. * math.pi)
Out[5]: 1.0

answered Jun 16 '15 at 16:11

Ami Tavory

74,578
11
141
185

1

Yes, I know... but the "atan" it's not supposed to give the angle in radians? – Bilis Jun 16 '15 at 16:19

score 0 · Answer 2 · edited May 23 '17 at 11:44

0

You should use np.array initialized to zero if you want to deal with big arrays (just a suggestion, it is working either way, i prefer to use arrays):

#from math import pi, sin,cos, atan 
import numpy as np 
a=1                 #float( input("Introduce distance: ")) 
N=10                #float( input("Introduce number of charges: ")) 
Q=10**-8 
s=np.linspace (0,1,N)
y=np.zeros(N)
x=np.zeros(N) 
z=np.zeros(N) 
ko=1/(4*np.pi*8.8542*10**-12)
for n in s: 
    f=np.arctan(n/a) 
    L=ko*(Q/(N*(a**2+n**2)))*np.cos(f) 
    k=ko*(Q/(N*(a**2+n**2)))*np.sin(f) 
    E=ko*(Q/(N*(a**2+n**2))) 
    z[n]=E 
    y[n]=k 
    x[n]=L 

zz=sum(z) 
yy=sum(y) 
xx=sum(x) 
print (x,y)       #xx
print(z)          #zz

The small difference that is there is because of float cannot hold more precise value. You can use np.pi/4 in place of f to get same values. If you want to use higher precision values, see this answer. If you want to simulate integration you might have to look into monte carlo simlations.

edited May 23 '17 at 11:44

Community

1
1

answered Jun 16 '15 at 16:42

DarthSpeedious

965
1
13
25

But im getting (63.551473451641151, 26.323789142728536) instead of (63.551500711322454, 63.55150071132246) that is what i'm supposed to get, so I don't think the problem is in the float... But I don't know... :/ – Bilis Jun 16 '15 at 17:15
Ok, so your way is actually functional, touché, the components are the same when the angle is 45º but the numbers aren't the ones that I get with the integral. – Bilis Jun 16 '15 at 17:47
Integral!? can you please explain, what integration are we talking about? Since your question said charges earlier did you mean continuous integral? – DarthSpeedious Jun 16 '15 at 20:29

Wrong cosine values in Python

2 Answers2