Python advanced 3.the python std lib by example – algorithm
THE PYTHON STD LIB BY
EXAMPLE – ALGORITHM
Saturday, December 21, 2013
• Python includes several modules which can
implement algorithm elegantly and concisely.
• It support uprely procedural, OOP and
• It includes: functools, partial , itertools,
operator, contextlib etc
Partial Objects: provide default
• The partial objects can provide or change the default
value of the argument.
• Example code (assume we have define function
>>> import functools
>>> p1 = functools.partial(myfunc,b=4)
>>> p1(‘passing a’)
>>> p2 = functools.partial(myfunc,’default a’,b=99)
>>> p2(b=‘override b’)
• The partial object does not have __name__
and __doc__ attributes by default.
• Using update_wrapper(0 copies or added
attributes from the original function.
the “rich comparison”
First let us learn which is “rich comparion” in
•Rich comparison method API (__lt__, __le__,
__eq__, __gt__, __ge__)
(Here le means less than, le means “less or equal”,
gt means “greater than”, ge means “greater than
•These method API can help perform a single
comparison operation and return a Boolean value
Example of rich comparision
We implement __eq__ and __gt__.
Functools.total_ordering can implement other
operator (<, <=, >= etc) base on eq and gt.
Function cmp_to_key: convert
cmp to key for sorting
• In Python 2.xx, cmp do comparion:
cmp(2,1) -> 1
cmp(1,1) -> 0
cmp(1,2) -> -1
• In python 3, cmp in sort function no longer
• Functools.cmp_to_key convert cmp to key
Quick example of cmp_to_key
• Built-in funtion cmp need two argument.
• Sorted function can use other option
key=func. Sorted by key (only support this on
• The itertools module includes a set of
functions for working with sequence data
sets (list, tuple,set,dict etc).
• Iterator based code offer better memory
Function chain(): Merge iterators
• Take serveral iterators as arguments and
return a single iterator
Function imap: similar as map
• Imap accept a function, and multiple
sequences, return a tuple.
Other function merge and split
• Function izip: like zip, but combine iterator
and return iterator of tuple instead of list
• Function islice: similar as slice
• Function imap: similar as map
• Function ifilter: similar as filter, filter those
items test functions return True
• Function ifilterfalse: filter those items where
the test function return False
• First, let us review the star * syntax in Python.
• Star * means unpack the sequence reference as
>>> def foo(bar,lee):
>>> a = [1,2]
>>> foo(a) # it is wrong, need two arguments
>>> foo(*a) # it is right. The list is unpack
>>>foo(1,2) # it is the same thing
Function starmap: unpack the
• Unpack the item as argument using the *
Function count(): iterator produce
• Function count(start=0,step=1): user can pass
the start and step value.No upper bound
>>> a = itertools.count(start=10,step=10)
>>> for i in a:
if I >100:
Print list 10,20,30 … 110
Function cycle: iterator do
• It need remember the whole input, so it may
consume quite a bit memo if input iterator is
Function repeat: repeat same
value several time
• This example mean repeat ‘a’ 5 times.
>>> itertools.repeat(‘a’, 5)
It is similar as list [‘a’,’a’.’a’,’a’,’a’]
The return is a iterator but not list. So it use the
memo only when it is called.
Function dropwhile and takewhile
• Func dropwhile start output while condition
become false for the first time
• Example, 3rd element do not met x<1. So it
return 3 to end of this list
Function dropwhile and takewhile
• The opposite of dropwhile: stop output while
condition become false for the first time
• So all output items meet the condition
A particular slide catching your eye?
Clipping is a handy way to collect important slides you want to go back to later.