University:
Boston University
Course:
MET CS 201 | Introduction to Programming
Academic year:

2021
Views:

141

Pages:

7
Author:

Lilian Dillon

Outline       Introduction Profiling Data Structures Syntax Generators Accelerators Python Syntax  Here are some common ways where Python syntax can result in unintended consequences. String concatenation  Avoid excessive use of ‘+’ as each ‘+’ creates a temporary string.  Very time and memory-intensive in loops. Best to avoid this # a list of some strings strs = ['a','b','c',...] s = "" for i in strs: s += i Less code, faster, and less memory.  If strings are in a list (or similar thing) use the string join() function.  Building a string in a loop? Append them to a list then join(). strs = ['a','b','c',...] s = "".join(strs) msg = [] for idx,elem in enumerate(some_data): # doing something... # record a message/result/etc msg.append('Step %s complete\n' % idx) # Now concatenate msg = ''.join(msg) Slice Notation  Lists:  RHS list slicing copies lists  LHS list slicing overwrites elements  Numpy ndarrays:  RHS ndarray slicing creates a Numpy view  LHS ndarray slicing overwrites elements x = [1,2,3,4] # y is a new list y = x[0:2] # y --> [1,2] x[0:2] = [-5,-6] # 1st two elements # of x are overwritten # x --> [-5,-6,3,4] x = numpy.array([1,2,3,4]) # y is a view into x y = x[0:2] # y --> x0:2] --> [1,2] x[0:2] = [-5,-6] # 1st two elements # of x are overwritten # x --> [-5,-6,3,4] # y --> x[0:2] --> [-5,-6] No. The del command  Temporary variables in loops – avoid the del command to clear out lists.  The del works by marking the elements of list x for deletion at some later time, not when the del is called.  The cleared elements of x aren’t cleaned up until x goes out of scope.  This can result in a surprising amount of memory consumption!  Instead re-declare x with each inner loop iteration. x=[] sum = 0.0 for i in range(N): for j in range(M): # do something that # adds stuff to x sum += sum(x) # clear out x del x[:] Yes. sum = for i x = for # # sum 0.0 in range(N): [] j in range(M): do something that adds stuff to x += sum(x) Open files with with  The Python with command when opening files will auto-handle the closing of the file.  The operating system limits the number of files that can be opened…it’s easy to forget a file .close() call. import glob import os files = glob.glob(os.path.join(img_dir,'*.dat')) # Do something with each data file for datfile in files: dat=open(datfile,'r') some_func(dat.read()) # If there are enough files and you # don't call this: # dat.close() # this loop WILL CRASH when you hit # your open file limit. # Life is better with "with" : for datfile in files: with open(datfile,'r') as dat: some_func(dat.read()) # this guarantees the open file is # closed when this code block ends dat.close() ## this is now optional. Python’s itertools and functools libraries  These two libraries are full of highly useful tools for manipulating Python functions and data structures.  Well worth checking out!  itertools:  “The module standardizes a core set of fast, memory efficient tools that are useful by themselves or in combination. Together, they form an “iterator algebra” making it possible to construct specialized tools succinctly and efficiently in pure Python.”  functools:  “The functools module is for higher-order functions: functions that act on or return other functions. In general, any callable object can be treated as a function for the purposes of this module.”