ipython-error "list indices must be integers" , they ARE integers..

Question

HI there,

I am having a problem. I have an array of 31 elements called colors. I also have another array with integer numbers that vary between 0 and 31, this is called c. I want to produce a new array where the values in c are now the corresponding values in colors.

I write:

newarray=colors[c]

but get the error message "list indices must be integers" but c is an array of integers. I am new to python and have not got time to do the tutorials as I just need it for a specific plotting task. Could anyone give me a hand?

Thanks

Answer 1

array of integers != integer

list indices must be integers - you've given a list of integers.

You probably want a list comprehension:

newarray = [ colors[i] for i in c ]

But you really ought to learn the language properly before using it. It's not like python takes a long time to learn.

EDIT:

If you're still getting the same error then your assertion that c is a list of integers is incorrect.

Please try:

print type(c)
print type(c[0])
print type(colors)
print type(colors[0])

Then we can work out what types you have got. Also a short but complete example would help, and probably teach you a lot about your problem.

EDIT2:

So if c is actually a list of string, you should probably have mentioned this, strings don't get automatically converted to integers, unlike some other scripting languages.

newarray = [ colors[int(i)] for i in c ]

EDIT3:

Here is some minimal code that demonstrates a couple of bug fixes:

x=["1\t2\t3","4\t5\t6","1\t2\t0","1\t2\t31"]
a=[y.split('\t')[0] for y in x]
b=[y.split('\t')[1] for y in x]
c=[y.split('\t')[2] for y in x]  # this line is probably the problem
colors=['#FFFF00']*32
newarray=[colors[int(i)] for i in c]
print newarray

a) colors needs to be 32 entries long. b) the elements from c (i) in the list comprehension need to be converted to integers (int(i)).

Answer 2

Python does not support arbitrary list indexing. You can use single integer index like c[4] or slice like c[4:10]. SciPy library has richer indexing capabilities. Or just use list comprehensions as Douglas Leeder advices.

Answer 3

Okay, I think I know what you are after...

So you have your list of 31 colours. Say for argument it is a list of 31 strings like this...

colours = [ "Black", "DarkBlue", "DarkGreen", ... "White" ]

And 'c' is an array of numbers in the range 0 to 31, but in random order...

import random
c = [x for x in xrange(32)]
random.shuffle(c)
# 'c' now has numbers 0 to 31 in random order
# c = [ 2, 31, 0, ... 1]

And what you want to do is map the values in c as an index into the list of colors so that you end up with a list of colors in as indexed by c and in that order...

mapped = [color[idx] for idx in c]
# mapped has the colors in the same order as indexed by 'c'
# mapped = ["DarkGreen", "White", "Black", ... "DarkBlue"]

If that ain't what you want then you need to revise your question!

I think you've got a basic problem in that your list of colours should have 32 elements (colours) in it, not 31, if the list 'c' is a random shuffle of all numbers in the range 0 to 31 (that's 32 numbers, you see).

Answer 4

This is your code: (from your comment)

from pylab import* 
f=open("transformdata.txt") 
x=f.readlines() 
a=[y.split('\t')[0] for y in x] 
b=[y.split('\t')[1] for y in x] 
c=[y.split('\t')[2] for y in x]  # this line is probably the problem
subplot(111,projection="hammer") 
colors=['#FFFF00']*31
newarray=[colors [i] for i in c] 
p=plot([a],[b],"o",mfc=color) show()

Without knowing exactly what your data is, or what you're trying to accomplish, I'd suggest trying this:

c=[int(y.split('\t')[2]) for y in x]