How To Calculate Centroid In Python

I'm beginner to python coding. I'm working over structural coordinates. I have pdb structure which have xyz coordinate information (last three col) ATOM 1 N SER A 1

Solution 1:

First of all, an easier way to read your file is with numpy's genfromtxt function. You don't need to import string, and you don't need to loop through all the lines and append text or count the characters.

from __future__ import division
import numpy as nx

data = nx.genfromtxt('file.pdb')

Then, the last three columns can be accessed as:

data[:, -3:]

Where the first : means "all rows", and -3: means from the third-to-last column to the last column.

So, you can average them as such:

nx.mean(data[:,-3:], axis=0)

where the axis=0 argument tells nx.mean to take the average along the first (0th) axis. It looks like this:

In : data[:,-3:]
Out: 
array([[ 27.13,   7.77,  34.39],
       [ 27.99,   7.76,  34.93],
       [ 27.16,   6.96,  33.79],
       [ 27.17,   8.58,  33.79],
       [ 25.94,   7.78,  35.25],
       [ 25.98,   9.09,  36.02],
       [ 26.74,  10.1 ,  35.32],
       [ 26.75,  10.94,  35.86],
       [ 24.64,   7.79,  34.46],
       [ 24.53,   8.51,  33.5 ],
       [ 23.59,   7.07,  34.76],
       [ 23.59,   6.55,  35.61],
       [ 22.42,   7.01,  33.9 ],
       [ 21.62,   5.76,  34.27],
       [ 22.48,   4.21,  33.97],
       [ 21.59,   8.22,  34.04],
       [ 21.37,   8.69,  35.16]])

In : np.mean(data[:,-3:], axis=0)
Out: array([ 24.74647059,   7.81117647,  34.64823529])

---

Some other things:

1) remove this line: import math as mean, which imports the entire math module and renames it mean. What you intended was from math import mean which imports the mean function from the math module. But in your code, you end up using the math function from the numpy (nx) module anyway, so you never used the math version.

2) your loop is not indented, which means you either pasted incorrectly into StackOverflow, or your loop is incorrectly indented. Possibly, this is what your code actually looks like:

for j in text:
    x1 = eval(replace(j[30:38], ' ', ''))         #extract x-coordinate
    y1 = eval(replace(j[38:46], ' ', ''))         #extract y-coordinate
    z1 = eval(replace(j[46:54], ' ', ''))         #extract z-coordinate

    idcord = []
    idcord.append(x1); idcord.append(y1); idcord.append(z1)

    centroid = nx.mean(idcord)
    print centroid

But the problem is that idcord gets set to an empty list every time the loop goes through, and a new centroid is calculated, for each particle. You don't even need the loop at all if you import the data file all at once as above. In fact, your entire code can be:

from __future__ import division
import numpy as nx

data = nx.genfromtxt('file.pdb')
nx.mean(data[:,-3:], axis=0)

Solution 2:

try this

import numpy as nx
X = nx.rand(10,3)   # generate some number
centroid = nx.mean(X)
print centroid