nDimensional sparse histogram in python.

[Kirk McDonald]

KraftDiner wrote:
> Ok so this is nice.. Just one thing.. When you try to get a value from
> a dictionary
> and it isn't found in the dictionary things go bad...
> 
> Take this for example:
> 
> class histogram(object):
> 	def __init__(self):
> 		self.histo = {}
> 
> 	def update(self, point):
> 		if self.histo.get(point) != None:
> 			self.histo[point] = self.histo[point] + 1
> 		else:
> 			self.histo[point] = 1
> 
> 	def get(self, point):
> 		return self.histo[point]
> 
> 
> hist = histogram()
> hist.update((0,0,0))
> hist.update((0,0,1))
> hist.update((0,0,1))
> hist.get((0,0,0))
> hist.get((0,0,1))
> hist.get((0,0,2))
> 
> spews out this error:
> 
> Traceback (most recent call last):
>   File "histogram.py", line 21, in ?
>     hist.get((0,0,2))
>   File "histogram.py", line 12, in get
>     return self.histo[point]
> KeyError: (0, 0, 2)
> 

Try this:

class histogram(object):
	def __init__(self):
		self.histo = {}

	def update(self, point):
		self.histo[point] = self.histo.get(point, 0) + 1

	def get(self, point):
		return self.histo.get(point, 0)

dict.get's default return value is your friend.

-Kirk McDonald