Ambiguous Evaluation of Lambda Expression on Array

Question

I would like to use a lambda that adds one to x if x is equal to zero. I have tried the following expressions:

t = map(lambda x: x+1 if x==0 else x, numpy.array())
t = map(lambda x: x==0 and x+1 or x, numpy.array())
t = numpy.apply_along_axis(lambda x: x+1 if x==0 else x, 0, numpy.array())

Each of these expressions returns the following error:

ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

My understanding of map() and numpy.apply_along_axis() was that it would take some function and apply it to each value of an array. From the error it seems that the the lambda is being evaluated as x=array, not some value in array. What am I doing wrong?

I know that I could write a function to accomplish this but I want to become more familiar with the functional programming aspects of python.

What is numpy.array? I assume it's not the numpy function of that name? What is the data you're trying to apply this to? — BrenBarn
– BrenBarn, Commented Nov 23, 2012 at 23:01
It works for me; for example, map(lambda x: x+1 if x==0 else x, np.array([0, 1, 2, 3])) evaluates to [1, 1, 2, 3], as expected. Please provide a minimal working example that exhibits the behavior you're describing. — user4815162342
– user4815162342, Commented Nov 23, 2012 at 23:03
On a side note, x+1 if x==0 can be written as 1 if x==0 or 1 if not x. x==0 and x+1 or x I find not very clear. Perhaps the shortest form for it all is x if x else 1. — Thijs van Dien
– Thijs van Dien, Commented Nov 23, 2012 at 23:09

ecatmur · Accepted Answer · 2012-11-23 23:03:38Z

8

If you're using numpy, you should be writing vectorised code:

arr + (arr == 0)

answered Nov 23, 2012 at 23:03

ecatmur

158k28 gold badges311 silver badges387 bronze badges

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

4 Comments

Joe Over a year ago

Could you explain this a bit more? I'm not sure what is going on here.

ecatmur Over a year ago

@Joe looping over every element of a large array in Python is slow. Arithmetic, logical and comparison operators on a numpy array are treated elementwise, so this will give you the result in a far more efficient form.

Joe Over a year ago

I get the goal of vectorization, I just don't exactly understand how the expression you posted works.

ecatmur Over a year ago

@Joe (arr == 0) gives a boolean array that is True where arr is equal to 0. In arithmetic, True is converted to 1 and False to 0, so this adds 1 where arr is equal to 0.

Collectives™ on Stack Overflow

Ambiguous Evaluation of Lambda Expression on Array

1 Answer 1

4 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

4 Comments

Your Answer

Sign up or log in

Post as a guest

Related