Eucledian distance to point source

user_14139020 0
Viewed 90

I am stimulating a model via a point source, which is located above (z-direction)-- to be able to compute the impact of the stimulation i need to compute the eucledian distance from this point power source to each mid of compartment (see picute).

I tried it this way, but the results are strange -- maybe the computation of the distance is wrong...

x_Mid = np.zeros(nComp)
y_Mid = np.zeros(nComp)
z_Mid = np.zeros(nComp)
for i in range(0, nComp):
    y_Mid[i] = 0.
    if i == 0:
        x_Mid[i] = (lComp[i] / 2.)
        z_Mid[i] = 1*elecShift
        compDist[i] = distance.euclidean(x_Mid,y_Mid,z_Mid)*10**(-4)

    else:
        x_Mid[i] = x_Mid[i - 1] + (lComp[i - 1] / 2.) + (lComp[i] / 2.)
        z_Mid[i] = 1*elecShift
        compDist[i] = distance.euclidean(x_Mid,y_Mid,z_Mid)*10**(-4)

lcomp is the length of the compartment. y - direction is zero, because its a 2D Model. elecshift is the distance of the point source in z-direction and the units are micrometer (therefore then its multiplied by 10^-4 to give it in centimeter). nComp is the number of compartments.

Is the computation of the eucledean distance from the source to each compartment center correct?

pic1

scipy numpy python
1 Answers

I'm assuming the source is at [0, 0, 0].

You can calculate three vectors in a simpler way:

x_Mid = np.cumsum(lComp) - lComp / 2.
y_Mid = np.zeros_like(x_Mid)
z_Mid = elecShift * np.ones_like(x_Mid)

Then the simplest calculation of distance is just:

compDist = np.sqrt(x_Mid**2 + y_Mid**2 + z_Mid**2) * 1.e-4

or even:

compDist = np.sqrt(x_Mid**2 + elecShift**2) * 1.e-4

And if you want to use function from scipy, then according to API, use:

for i in range(0, nComp):
  compDist[i] = distance.euclidean([x_Mid[i], y_Mid[i], z_Mid[i]], 0.)*10**(-4)

Your code was providing current x_Mid as one point, y_Mid as second one and z_Mid as weights to distance.euclidean().

user_133660240
Related