Python, load and list a .csv in a numpy array

Question

I am loading a .csv file with this code:

import csv
import numpy as np
import scipy.spatial

points = np.array([((int(R), int(G), int(B)),float(X), float(Y), float(Z))
              for R, G, B, X, Y, Z in csv.reader(open('XYZcolorlist_D65.csv'))])
    # load R,G,B,X,Y,Z coordinates of 'points' in a np.array  

print points

And that works fine.

However, if I add this further line, where I am trying to compute a Delaunay triangulation with scipy: http://docs.scipy.org/doc/scipy/reference/generated/scipy.spatial.Delaunay.html

tri = scipy.spatial.Delaunay(points[1, 2, 3])
# do the triangulation

I get this error message:

Traceback (most recent call last):
  File "C:\Users\gary\Documents\EPSON STUDIES\delaunay.py", line 15, in <module>
    tri = scipy.spatial.Delaunay(points[1, 2, 3])
IndexError: too many indices

Obviously the syntax scipy.spatial.Delaunay(points[1, 2, 3]) isn't correct.

What is it that I am doing wrong?

EDIT:

If it's easier to handle, I can also use this line to import (np arrays should be of same type of data?)

points = np.array([(float(R), float(G), float(B), float(X), float(Y), float(Z))
              for R, G, B, X, Y, Z in csv.reader(open('XYZcolorlist_D65.csv'))])
    # load R,G,B,X,Y,Z coordinates of 'points' in a np.array

Then I would need to skip the first 3 values in each row...

map(float, row) for row in reader is simpler syntax.

cohoz
– cohoz

2014-02-07 03:39:31 +00:00
Commented Feb 7, 2014 at 3:39 — cohoz
– cohoz, Commented Feb 7, 2014 at 3:39

YS-L · Accepted Answer · 2014-02-07 04:14:20Z

1

The error is with the slicing of numpy array. To get the coordinates of the points, either version of you code is fine.

First version of your code, where the first column of points is a tuple of RGB values:

tri = scipy.spatial.Delaunay(points[:, 1:])

Second version, where the RGB values are flattened, taking up 3 columns, you need to skip three columns:

tri = scipy.spatial.Delaunay(points[:, 3:])

In fact, you can use np.loadtxt for reading in the data (yielding the second version):

points = np.loadtxt('XYZcolorlist_D65.csv')

edited Feb 7, 2014 at 4:14

answered Feb 7, 2014 at 2:30

YS-L

14.8k4 gold badges52 silver badges62 bronze badges

Sign up to request clarification or add additional context in comments.

5 Comments

YS-L Over a year ago

Mind posting first few row of points? What is the dimension of the points that needs to be triangulated?

adrienlucca.net Over a year ago

print points[1:4] returns the first 3 rows:

[[(255, 95, 127) 40.2074945517 26.5282949405 22.7094284437]  [(255, 127, 127) 43.6647438365 32.3482625492 23.6181801523]  [(255, 159, 127) 47.1225628354 39.1780944388 22.9366615044]]

, I want to get the values inside, not the entire rows

YS-L Over a year ago

get the values inside - can you give an example? For the first point, you want [40.2, 26.5, 22.7]?

YS-L Over a year ago

Do you actually want all the points in points, but skipping the first column which represents the color?

adrienlucca.net Over a year ago

EDIT: I added another way of loading the csv

cohoz · Accepted Answer · 2014-02-07 04:04:36Z

0

I'm not sure exactly what you're trying to do, but maybe it's something like this...you can replace file_txt with your csv reader.

>>> from scipy.spatial import Delaunay
>>> file_txt = [[255, 95, 127, 40.2, 26.5, 22.7], [255, 127, 127, 43.6, 32.3, 23.6], [255, 159, 127, 47.1, 39.1, 22.9], [0,0,0,0,0,0]]
>>> a = [list(map(float, row[3:])) for row in file_txt]
>>> a
[[40.2, 26.5, 22.7], [43.6, 32.3, 23.6], [47.1, 39.1, 22.9], [0.0, 0.0, 0.0]]
>>> Delaunay(a)
<scipy.spatial.qhull.Delaunay object at 0x0383FA30>

answered Feb 7, 2014 at 4:04

cohoz

7606 silver badges16 bronze badges

Collectives™ on Stack Overflow

Python, load and list a .csv in a numpy array

2 Answers 2

5 Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

5 Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related