Tuple packing and unpacking in Python
I was chatting to a mate who has recently started learning Python. He mentioned that one of the cool features that got him sold on Python is the fact that you can swap two values with the following idiom:
x, y = y, x
Yes, that is neat and quite predictable given that Python has automatic packing and unpacking of tuples. In other words, Python packs the values on the right hand side into a tuple and the assignment unpacks each value on the right hand side into the values on the left hand side.
Extending this, we can do something like this:
a = [3, 1, 4]
x, y, z = sorted(a)
This will give us the lowest value in x and the highest value in z.
Once again, this is a nice little hack. I was trying to think of an example where packing and unpacking really adds some value (in terms of readability or ‘Pythonicness’). This morning, I accidentally came across the following example:
Suppose we want to generate a sequence of Fibonacci numbers (anyone who has done CS101 will recall that Fibonacci Numbers are the sequence of numbers 0, 1, 1, 2, 3, 5, 8, 13… where each number is created from the sum of the previous two numbers). Fibonacci, who went by many names, was apparently trying to calculate how many rabbits he would have if he started with one parent and each parent miraculously reproduced every 6 months having exactly one baby. After 12 months that baby would become a parent. Of course these miracle bunnies never die.
The following program calculates the rabbit population until before it hits 1,000:
``` {.sourceCode .python} parent, baby = 1, 0
while baby < 1000: print baby parent, baby = (parent + baby, parent) ```
I think this is a wonderful example of how one of Python’s language features makes code really clear and how Python earns its reputation of ‘executable pseudo-code’.
Postscript: Another nice little way we can use this feature is as follows:
day, month, year = '31/12/2010'.split('/')