is_point

[calvinmetcalf]

@scardine your function

def is_point(p):
    try:
        float(p[0]), float(p[1])
    except (TypeError, IndexError):
        return False

is not equivalent to

is_point = lambda x : type(x)==type([]) and len(x)==2

[scardine]

No, it isn't because it will not return false if len(p) > 2.

Python programmers usually write isinstance(p, list) instead oftype(a) == type([])`, because type([]) will trigger memory allocation for an empty list. Better yet is not check for an specific type, but to use duck-typing/EAFP: just carry out the operation and catch relevant exceptions.

This will check if p is an iterable of numbers or numeric strings elements with length 2:

def is_point(p):
    try:
        return len([float(i) for i in p[:4]]) == 2
    except (TypeError, IndexError):
        return False

The slicing in p[:4] is to avoid checking if all members of p are numeric if p is larger than 3.

[calvinmetcalf]

what about

def is_point(p):
    try:
        return len(p)==2
    except (TypeError, IndexError):
        return False

[scardine]

>> is_point('Ha')
True

>> is_point(['something', 'else'])
True

Previous suggestion was checking:

• if the argument is an iterable of length 2
• if argument[0] and argument[1] are int, float or a numeric string

[calvinmetcalf]

I'm going to keep it as it is until we get tests working I think:

def is_point(p):
    try:
        return len([float(i) for i in p[:4]]) == 2
    except (TypeError, IndexError):
        return False

looks good, but i really want test coverage to make sure.