f ilter vs comprehension-if

Le Slid
ctu e
re 33
Lesson #4: The filter may save computation:

10
>>> from x import prime
>>> [2*x+1 for x in range(15) if prime (2*x+1)]
[3, 5, 7, 11, 13, 17, 19, 23, 29]
>>> list(filter(prime,[2*x+1 for x in range(15)]))
[3, 5, 7, 11, 13, 17, 19, 23, 29]
Later we’ll learn about the
Notice that the comprehension required
“assignment the
expression.”
“2*x+1” to be typed twice. That allows this computation
to only be performed
This also means that it is calculated twice. once.
If “2*x+1” were replaced with something more
complex (eg: sqrt(sqrt(sqrt(x)))), then the cost of
calculating twice could be significant for long lists.
 Expression Assignments
>>> from random import randrange as Rn
>>> Ns=range(10) What this output means is that 6
of the randomiterations
>>> [x for x in Ns if Rn(9)<5]#Which numbers were <
match?
[0, 2, 6] 5.
For each of those,
>>> [x for x in Ns if Rn(9)<5]#Which a new random
iterations match?
[1, 2, 3, 5, 7] number was generated. Some of
those new
>>> #But how to get those values? numbers
This won'tare  5.
work:
>>> [R(9) for x in Ns if Rn(9)<5]#This gives #s >= 5
[7, 2, 7, 0, 8, 5]
>>> #The problem is the expression value became lost.
>>> #Use an "expression assignment" to keep it:
>>> [y for x in Ns if (y:=Rn(9))<5]
[0, 1, 3, 1]
>>> [y for x in Ns if (y:=Rn(9)<5)]#Saves the < exprsn
[True, True, True]
>>> [y for x in Ns if y:=Rn(9)<5] # We needed the ( )
 Expression Assignments
>>> from random import randrange as Rn
>>> Ns=range(10)
>>> [x for x in Ns if Rn(9)<5]#Which iterations match?
[0, 2, 6]
>>> [x for x in Ns if Rn(9)<5]#Which iterations match?
[1, 2, 3, 5, 7]
>>> #But how to get those values? This won't work:
>>> [R(9) for x in Ns if Rn(9)<5]#This gives #s >= 5
[7, 2, 7, 0, 8, 5]
>>> #The problem is the expression value became lost.
>>> #Use an "expression assignment" to keep it:
>>> [y for x in Ns if (y:=Rn(9))<5]
[0, 1, 3, 1]
>>> [y for x in Ns if (y:=Rn(9)<5)]#Saves the < exprsn
[True, True, True]
>>> [y for x in Ns if y:=Rn(9)<5] # We needed the ( )
File "<stdin>", line 1
 Expression Assignments
>>> from random import randrange as Rn
>>> Ns=range(10)
>>> [x for x in Ns if Rn(9)<5]#Which iterations match?
[0, 2, 6]
>>> [x for x in Ns if Rn(9)<5]#Which iterations match?
[1, 2, 3, 5, 7]
>>> #But how to get those values? This won't work:
>>> [R(9) for x in Ns if Rn(9)<5]#This gives #s >= 5
[7, 2, 7, 0, 8, 5]
>>> #The problem is the expression value became lost.
>>> #Use an "expression assignment" to keep it:
>>> [y for x in Ns if (y:=Rn(9))<5]
[0, 1, 3, 1]
>>> [y for x in Ns if (y:=Rn(9)<5)]#Saves the < exprsn
[True, True, True]
>>> [y for x in Ns if y:=Rn(9)<5] # We needed the ( )
File "<stdin>", line 1
[y for x in Ns if y:=Rn(9)<5] # We needed the ( )
 Expression Assignments
>>> from random import randrange as Rn
>>> Ns=range(10)
>>> [x for x in Ns if Rn(9)<5]#Which iterations match?
[0, 2, 6]
>>> [x for x in Ns if Rn(9)<5]#Which iterations match?
[1, 2, 3, 5, 7]
>>> #But how to get those values? This won't work:
>>> [R(9) for x in Ns if Rn(9)<5]#This gives #s >= 5
[7, 2, 7, 0, 8, 5]
>>> #The problem is the expression value became lost.
>>> #Use an "expression assignment" to keep it:
>>> [y for x in Ns if (y:=Rn(9))<5]
[0, 1, 3, 1]
>>> [y for x in Ns if (y:=Rn(9)<5)]#Saves the < exprsn
[True, True, True]
>>> [y for x in Ns if y:=Rn(9)<5] # We needed the ( )
File "<stdin>", line 1
[y for x in Ns if y:=Rn(9)<5] # We needed the ( )
 Expression Assignments
>>> Ns=range(10)
>>> [x for x in Ns if Rn(9)<5]#Which iterations match?
[0, 2, 6]
>>> [x for x in Ns if Rn(9)<5]#Which iterations match?
[1, 2, 3, 5, 7]
>>> #But how to get those values? This won't work:
>>> [R(9) for x in Ns if Rn(9)<5]#This gives #s >= 5
[6, 2, 6, 0, 8, 5]
>>> #The problem is the expression value became lost.
>>> #Use an "expression assignment" to keep it:
>>> [y for x in Ns if (y:=Rn(9))<5]
[0, 1, 3, 1]
>>> [y for x in Ns if (y:=Rn(9)<5)]#Saves the < exprsn
[True, True, True]
>>> [y for x in Ns if y:=Rn(9)<5] # We needed the ( )
File "<stdin>", line 1
[y for x in Ns if y:=Rn(9)<5] # We needed the ( )
^
SyntaxError: invalid syntax
 Expression Assignments
>>> [x for x in Ns if Rn(9)<5]#Which iterations match?
[0, 2, 6] Works with no ( )
>>> [x for x in Ns if Rn(9)<5]#Which iterations match?
[1, 2, 3, 5, 7]
>>> #But how to get those values? This won't work:
>>> [R(9) for x in Ns if Rn(9)<5]#This gives #s >= 5
[7, 2, 7, 0, 8, 5]
Works with ( )
>>> #The problem is the expression value became lost.
>>> #Use an "expression assignment" to keep it:
>>> [y for x in Ns if (y:=Rn(9))<5]
[0, 1, 3, 1] Fails with no ( )
>>> [y for x in Ns if (y:=Rn(9)<5)]#Saves the < exprsn
[True, True, True]
>>> [y for x in Ns if y:=Rn(9)<5] # We needed the ( )
File "<stdin>", line 1
[y for x in Ns if y:=Rn(9)<5] # We needed the ( )
^
SyntaxError: invalid syntax
>>>
 Let's def ine a timer function
>>> from time import time, sleep
>>> def timer(com="START",was=[0]):
... ..if com=="START": was[0]=time()
... ..else: print(' ',
... .. . f'{int(1000000*(time()-was[0])):,d}','usec')
...
>>> timer(); sleep(2); timer("STOP")
2,000,056 usec
>>> def f(x): sleep(1); return x
...
>>> timer(); f(None); timer("STOP")
1,000,107 usec
>>> timer(); f(None); timer("STOP")
1,000,212 usec
>>> import os;os.system("cat timer.py")#Can download
this
from time import time
def timer(com="START",was=[0]):
if com=="START": was[0]=time()
else: print(' ',
 Expression Assignments
Can Save Time:
>>> from timer import timer; from time import sleep
>>> timer(); sleep(2); timer("STOP")
2,000,058 usec
>>> def f(x): sleep(1); return x
...
>>> timer(); f(None); timer("STOP")
1,000,126 usec
>>> from prime import prime #You can download this
>>> timer();[2*f(x)+1 for x in range(15)
... if prime(2*f(x)+1)]; timer("STOP")
[3, 5, 7, 11, 13, 17, 19, 23, 29]
24,036,369 usec
>>> timer();[y for x in range(15)
... if prime(y:=2*f(x)+1)]; timer("STOP")
[1, 3, 5, 7, 11, 13, 17, 19, 23, 29]
15,030,192 usec
>>> # It was 9 sec faster without the recompute
 Functional Programming
• Python allows functional programming style.
• But that is not to say that everyone should
use a functional programming style.
• Most Python programming is not functional.
• Yet it has some advantages (fewer bugs,
occasionally clearer code), so it can be
used sometimes, even in programs that
are predominantly procedural.
• Three Python functions are key:
• map( ), filter( ), functools.reduce()
 Different types of programming languages:
 Procedural languages like C let their functions/procedure
have side effects (eg, printing inside a function).
“function”).
 Declarative languages like SQL are used for databases.
You write a description
description ofof the
theproblem
problemtotobe besolved,
solved,and
and
the computer figures out how to perform the the computation.
 Object-oriented languages like Java manipulate
collections of objects.
objects. Objects
Objectshavehaveananinternal
internalstate
statewith
with
methods that
methods that query
query or
or modify
modify this
this state.
state.
C++ and Python
Python are
are languages
languages thatthatsupport
supportOOP, OOP,but but
don’t force the use of object-oriented features.
 In functional programming languages like ML everything is
a function
function.Function
Functionbodies
bodieshave
haveonly
onlyaareturn
returnstatement
statement
with no side effects. Computation is achieved by nesting
calls of calls of calls of functions.
In Python everything is an object, even functions. So, in
many cases, most things can be functions. Python doesn’t
require functional programming, but it supports its use.
>>>
Comparing filter() and map()
print(filter.__doc__[:54]+filter.__doc__[88:
])
filter(function or None, iterable) --> filter
object
items of iterable for which function(item) is
true. If function is None, return the items
that are true.
>>> print(map.__doc__[:38]+map.__doc__[60:])
map(func, *iterables) --> map object

computes the function using arguments from each

of the
>>>
iterables. Stops when the shortest iterable is
exhausted.
A mapping is a concept from math

Fr ctur
Le
om e
Your typical mathematical function is written as: y = f(x)

5
But what is f( )? Any function. Perhaps sin(). In that case,
sin(x)
we write: y = sin(x).
Here the sine function maps
maps aa value,
value, x,
x, to
to aa new
new value,
value, y.y.
For example: y=sin(0)  y=0,
if x=90, then  y=1.  y=1.
or y=sin(90)
y=sin(90)
You can look at an x value, move your finger vertically to
intersect the line, then horizontally to find the mapped-to
value.
A mapping is a concept from math

Fr ctur
Le
om e
sin(90)=1 is a mapping for just one value.

5
But what is the mapping, for all values?
That would be the set of all mappings.
A mapping is a concept from math

Fr ctur
Le
om e
sin(90)=1 is a mapping for just one value.

5
But what is the mapping, for all values?
That would be the set of all mappings.
Think
Think of
of it
it as
as aa look-up
look-up table:
table: 0 → 0.00000
10 → 0.17365
Q:How to make a table like this? 20 → 0.34202
30 → 0.50000
A:With map() 40 → 0.64279
... → ...
A mapping is a concept from math

Fr ctur
Le
om e
Think of it as a look-up table: 0 → 0.00000

5
10 → 0.17365
Q:How to make a table like this? 20 → 0.34202
30 → 0.50000
A:With map() 40 → 0.64279
Think of it as a look-up table: ... ...
0 →→ 0.00000
The sin( ) function takes one number and10returns
→ 0.17365
another one
Q:How to make
number. One a table
number. So it like this?
cannot 20 → a0.34202
understand list:
>>> from math import sin, radians 30 → 0.50000
A:With map()
>>> sin([0,radians(10), 40 → 0.64279
radians(20)])
Traceback (most recent call last): ... → ...
File "<stdin>", line 1, in <module>
TypeError: must be real number, not list
>>> # Use map() to apply a mapping to a list:
>>> map(sin,[0,radians(10), radians(20)])
[0.0, 0.17364817766693033, 0.3420201433256687]
A mapping is a concept from math

Fr ctur
Le
om e
Think of it as a look-up table: 0 → 0.00000

5
10 → 0.17365
Q:How to make a table like this? 20 → 0.34202
30 → 0.50000
A:With map() 40 → 0.64279
... → ...
>>> # Use map() to apply a mapping to a list:
>>> map(sin,[0,radians(10), radians(20)])
[0.0, 0.17364817766693033, 0.3420201433256687]
>>> #Can you ever have too many maps(), really?
>>> dict(zip(range(0,91,10),map(lambda
x:x/100000, ... map(round, map(lambda x:
\
100000*x,(map(sin, ... map(radians,
range(0,91,10)))))))))
>>> # Use map() to apply a mapping to a list:
{0: 0.0, 10: 0.17365, 20: 0.34202, 30: 0.5, 40:
>>> map(sin,[0,radians(10), radians(20)])
0.64279, 50: 0.76604, 60: 0.86603, 70: 0.93969,
[0.0, 0.17364817766693033, 0.3420201433256687]
 map applies a given function to each
>>>
element
from math import factorial, gamma
>>> factorial(range(0,10)) #This won’t work
>>>
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
Theismath
TypeError:an integer requiredIt (got type
also has therange)
""
>>> list(map(factorial,range(0,10)))#map
>>> module has fixes it
(gamma) function.
[1, 1, 2, 6, 24, 120, 720, 5040, 40320, 362880]
the "n!" This relates to
>>> gamma(range(1,11))#This generalizes
>>> factorial
(factorial
Traceback (most recent call last): factorial by the
File "<stdin>", line階乘1,) in <module> formula:
function.
TypeError: must be real number, not n! range
= (n + 1)
>>> list(map(gamma,range(1,11)))#But it needs map
>>>
[1.0, 1.0, 2.0, 6.0, 24.0, 120.0, 720.0, 5040.0,
40320.0, 362880.0]
>>> list(map(int,map(gamma,range(1,11))))#cleaner
>>>
[1, 1, 2, 6, 24, 120, 720, 5040, 40320, 362880]
>>>
 A map/lamda multi-list example:
 map() can be applied to more than one list.
– The lists have to have the same length.
– The meaning is f(x,y) or f(x,y,z), etc.
>>> a=[1,2,3];b=[10,20,30];c=[100,200,300]
>>> list(map(lambda x,y:x+y, a,b))
[11, 22, 33]
>>> list(map(lambda x,y,z:x+y+z, a,b,c))
[111, 222, 333]
>>> list(map(lambda x,y,z:x+y-z, a,b,c))
[-89, -178, -267]
>>>
 Another map( ) example:
% cat test.py
def fahrenheit(T): # ( 華氏度 )
return ((float(9)/5)*T + 32)
def celsius(T): # ( 攝氏溫度 )
return (float(5)/9)*(T-32)
temp = (36.5, 37, 37.5,39)
F = map(fahrenheit, temp); print(F)
F = list(F); print(F)
C = list(map(celsius, F)); print(C)
% python3 test.py
<map object at 0x6ffffc8be48>
[97.7, 98.60000000000001, 99.5, 102.2]
[36.5, 37.00000000000001, 37.5, 39.0]
%
 A map( ) example with lambda
By using lambda, we wouldn't have had to
create the functions fahrenheit() and celsius() :
>>> c = [39.2, 36.5, 37.3, 37.8]
>>> f = map(lambda x:(float(9)/5)*x+ 32,c)
>>> f=[*f]; f
[102.56, 97.7, 99.14, 100.03999999999999]
>>>
>>> c2= map(lambda x:(float(5)/9)*(x-32),f)
>>> [*c2]
[39.2, 36.5, 37.300000000000004, 37.8]
>>>
 Core functional programming tools
 map(function, sequence)
– Apply the function to each item in the sequence.
– For multi-value functions, give multiple sequences.
– Return sequence of values returned by the function.
 filter(function, sequence)
– Apply the function to each item in the sequence.
– The function's result is treated as a boolean.
– Return those items that pass the boolean filter.
 functools.reduce(function, sequence)
– Apply the function to each item in the sequence.
– The function must take 2 arguments & return 1 value.
– That returned value then becomes the first argument to the
function on the next iteration (the second argument is the
next item in the sequence).
– Return 1 value, the final value returned by the function.
 map(function, sequence(s))
– Apply the function to each item in the sequence.
– For multi-value functions, give multiple sequences.
– Return sequence of values returned by the function.

>>> def star(x,y): return x*y

...
>>> list(map(star, "ABCDEFGHI", range(5)))
['', 'B', 'CC', 'DDD', 'EEEE']
>>> print (map.__doc__)
map(func, *iterables) --> map object

Make an iterator that computes the function

using arguments from each of the iterables.
Stops when the shortest iterable is exhausted.
>>>
 filter(function, sequence)
– Apply the function to each item in the sequence.
– The function's result is treated as a boolean.
– Return those items that pass the boolean filter.

>>> def even(x): return ((x%2 == 0)

...
>>> list(filter(even, range(10)))
[0,2,4,6,8]
>>> print (print (filter.__doc__)
filter(function or None, iterable) --> filter
object

Return an iterator yielding those items of iterable

for which function(item) is true. If function is
None, return the items that are true.
>>>
 functools.reduce(function, sequence)
– Apply the function to each item in the sequence.
– The function must take 2 arguments & return 1 value.
– That returned value becomes the 1st argument to the function
on the next iteration (2nd argument is the next sequence item).
– Return 1 value, the final value returned by the function.
>>> from functools import reduce
>>> reduce(lambda x,y: x+y, range(10))
45
>>> reduce(lambda x,y: x+y, map(str,range(10)))
'0123456789'
>>> print(reduce.__doc__[:46]+reduce.__doc__[47:])
reduce(function, sequence [, initial]) -> value
Apply a function of 2 arguments cumulatively to the
items of a sequence, from left to right, so as to reduce
the sequence to a single value. For example,
reduce(lambda x, y: x+y, [1, 2, 3, 4, 5]) calculates
((((1+2)+3)+4)+5). If initial is present, it is placed
before the items of the sequence in the calculation, and
Q:Other
Why not reduce
define a + lambda
function examples
instead?
>>> def f(a,b)
Remember this return a if (a>b) else b
syntax?
Calculating the sum of numbers from 1 to 100:
A:isn't
It Yes,Functional
that’s better style. Use lambdas
Programming, per se (only
本身 for
). But
>>> reduce(lambda x, y: x+y, range(1,101))
functions
it is often that don’t
used in need names
functional because they are
programming,
5050
only referenced
because it avoids once.
an (The only
if/else reason
statement – that ahere
and
>>> sum(range(1,101)) #Could’ve used sum()
this
purely
5050
lambda is
functional named,
programis so as
wouldto make
only the
havelater
reduce
function
>>>
command
calls and easier
returns for students
(thus: no ifto read.)
statements).
Determining the maximum of a list of numerical
values
See, by But using
wait. reduce:
Luckily, But this version
>>>
sum()f sum()
= lambda
only a,b: a if (a
this was would> b)
workelse
on b
>>>
also reduce(f,
works on [47,11,42,102,13])
a number strings too. Eg
works.
102 numbers. sequence. reduce(…, ['H', 'i'])
>>>
>>> max([47,11,42,102,13])#could’ve used
max
A very complex functional program
% cat functionalExample.py
import random, functools
grammar=dict(S=[['NP','VP']],NP=[['Art','N']],
VP=[['V','NP']],Art=['the','a'],N=['boy','ball',
'girl','table'],V=['hit','heard','saw','liked'])
def gen(phrase):return isinstance(phrase,list)
and mend(gen, phrase) or (phrase in grammar) and
gen(random.choice(grammar[phrase])) or [phrase]
def mend(fn, args): return [item for res in
map(fn, args) for item in res]
print(functools.reduce(lambda x,y: x+" "+y,
gen('S')).capitalize()+".")
% python3 functionalExample.py
A ball hit the ball.
% python3 functionalExample.py
The boy liked the table.
%
A very complex functional program
% cat functionalExample.py
import random, functools
grammar=dict(S=[['NP','VP']],NP=[['Art','N']],
VP=[['V','NP']],Art=['the','a'],N=['boy','ball',
'girl','table'],V=['hit','heard','saw','liked'])
def gen(phrase):return isinstance(phrase,list)
and mend(gen, phrase) or (phrase in grammar) and
gen(random.choice(grammar[phrase])) or [phrase]
def mend(fn, args): return [item for res in
map(fn, args) for item in res]
print(functools.reduce(lambda x,y: x+" "+y,
gen('S')).capitalize()+".")
%Note: TherefunctionalExample.py
python3 are 2 functions (gen & mend) plus the top-level
A(ie, __main__).
ball hit theIgnoring
ball. definitions, there are only three
%statements:
python3 functionalExample.py
two returns and one print. That is how functional
The boy liked the table.
%programming works: the only statements are call and return.
A very complex functional program
% cat functionalExample.py
import random, functools
grammar=dict(S=[['NP','VP']],NP=[['Art','N']],
VP=[['V','NP']],Art=['the','a'],N=['boy','ball',
'girl','table'],V=['hit','heard','saw','liked'])
def gen(phrase):return isinstance(phrase,list)
and mend(gen, phrase) or (phrase in grammar) and
gen(random.choice(grammar[phrase])) or [phrase]
def mend(fn, args): return [item for res in
map(fn, args) for item in res]
print(functools.reduce(lambda x,y: x+" "+y,
gen('S')).capitalize()+".")
%Note: TherefunctionalExample.py
python3 are 2 functions (gen & mend) plus the top-level
A(ie, __main__).
ball hit theIgnoring
ball. definitions, there are only three
%statements:
python3 functionalExample.py
two returns and one print. That is how functional
The boy liked the table.
%programming works: the only statements are call and return.
A very complex functional program
% cat functionalExampleNoReturn.py
import random, functools
grammar=dict(S=[['NP','VP']],NP=[['Art','N']],
VP=[['V','NP']],Art=['the','a'],N=['boy','ball',
'girl','table'],V=['hit','heard','saw','liked'])
gen=lambda phrase: isinstance(phrase,list)
and mend(gen, phrase) or (phrase in grammar) and
gen(random.choice(grammar[phrase])) or [phrase]
mend=lambda fn, args: [item for res in
map(fn, args) for item in res]
print(functools.reduce(lambda x,y: x+" "+y,
gen('S')).capitalize()+".")
In fact, this
% python3 new version, functionalExampleNoReturn.py,
functionalExampleNoReturn.py
A ball hitdoesn’t even have return statements…
the girl.
% python3 functionalExampleNoReturn.py
The boy saw the boy.
%
>>> import mylib # (I wrote this module)
>>> mylib.displayfile("functionalExample.py")
from random import choice;from functools import
reduce
grammar=dict(S=[['NP','VP']],NP=[['Art','N']],
This method can
VP=[['V','NP']],Art=['the','a'],N=['boy','ball'
, colorfully display
a file’s contents.
'girl','table'],V=['hit','heard','saw','liked']
)

gen=lambda phrase: isinstance(phrase,list)\

and ((lambda fn,args:[el for r in map(fn,args)\
for el in r])(gen,phrase))or(phrase in
grammar)\
>>>
and gen(choice(grammar[phrase])) or [phrase]
I googled for an example of nontrivial
print(reduce(lambda x,y:
Python code written
x+' in the functional style,
'+y,gen('S')).capitalize()
and I found the code that I’ll now show... v
+'.')
 Well,#yes
>>> import mylib this is this
(I wrote a functional
module) example,
because it only has call statements.
>>> mylib.displayfile("functionalExample.py")
from random import choice;from functools import
reduce But it’s not purely functional, because
that would require it to all go onto one
grammar=dict(S=[['NP','VP']],NP=[['Art','N']],
VP=[['V','NP']],Art=['the','a'],N=['boy','ball'
,
line, with no variables (i.e., no “=”).
'girl','table'],V=['hit','heard','saw','liked']
So I tried rewriting it
)
in such a way as to
removeisinstance(phrase,list)\
gen=lambda phrase: this “=”.
and ((lambda fn,args:[el for r in map(fn,args)\
But errors kept
for el in r])(gen,phrase))or(phrase in
grammar)\ coming up, due to
>>>
and gen(choice(grammar[phrase])) or [phrase]
issues with this
I googled for an example of nontrivial
recursive
print(reduce(lambda
After 90 call.
x,y:
min of failing,
Python code written in the functional style,
x+' '+y,gen('S')).capitalize()
+'.') I thought
and I found the codeof ChatGPT.
that I’ve now shown.
I have the following Python code:
gen=lambda phrase: isinstance(phrase,list) and
((lambda fn, args: [item for res in map(fn, args) for item
in res])(gen, phrase)) or (phrase in grammar) and
gen(random.choice(grammar[phrase])) or [phrase]
Give me a recursive version that can be anonymous.

Sure, here's an anonymous recursive version of the

same function:
(lambda gen: (lambda f: f(f))(lambda f: lambda phrase:
isinstance(phrase, list) and sum(map(f, phrase), []) or
phrase in grammar and
gen(random.choice(grammar[phrase])) or [phrase]))
(lambda phrase: isinstance(phrase, list) and
sum(map(gen, phrase), []) or phrase in grammar and
I have the following Python code:
gen=lambda phrase: isinstance(phrase,list) and
((lambda fn, args: [item for res in map(fn, args) for item
in res])(gen, phrase)) or (phrase in grammar) and
gen(random.choice(grammar[phrase])) or [phrase]
Give me a recursive version that can be anonymous.

Sure, here's an anonymous recursive version of the

same function:
(lambda gen: (lambda f: f(f))(lambda f: lambda phrase:
isinstance(phrase, list) and sum(map(f, phrase), []) or
phrase in grammar and
gen(random.choice(grammar[phrase])) or [phrase]))
(lambda phrase: isinstance(phrase, list) and
sum(map(gen, phrase), []) or phrase in grammar and
Wait. Your code isn't quite what I want. I want to be able
to call it with only one line. See, this works in two lines:
>>> gen=lambda phrase: isinstance(phrase,list) and
((lambda fn, args: [item for res in map(fn, args) for item
in res])(gen, phrase)) or (phrase in grammar) and
gen(random.choice(grammar[phrase])) or [phrase]
>>> gen('S')
['a', 'ball', 'liked', 'the', 'table']
But I want to do it in one line:
>>> (lambda .....)('S')

Here's the modified code to achieve that:

(lambda phrase, g: g(phrase, g))('S', lambda phrase, g:
isinstance(phrase,list) and ((lambda fn, args: [item for
res in map(fn, args) for item in res])(g, phrase)) or
Wait. Your code isn't quite what I want. I want to be able
to call it with only one line. See, this works in two lines:
>>> gen=lambda phrase: isinstance(phrase,list) and
((lambda fn, args: [item for res in map(fn, args) for item
in res])(gen, phrase)) or (phrase in grammar) and
gen(random.choice(grammar[phrase])) or [phrase]
>>> gen('S')
['a', 'ball', 'liked', 'the', 'table']
But I want to do it in one line:
>>> (lambda .....)('S')

Here's the modified code to achieve that:

(lambda phrase, g: g(phrase, g))('S', lambda phrase, g:
isinstance(phrase,list) and ((lambda fn, args: [item for
res in map(fn, args) for item in res])(g, phrase)) or
When I ran your code, I got this error:
TypeError: <lambda>() missing 1 required positional
argument: 'g'
I apologize for the error in my previous response. Here's
a corrected version that should work in a single line:
(lambda f: (lambda x: x(x))(lambda y: f(lambda *args: y(y)
( *args))))(lambda g: lambda phrase: isinstance(phrase,
list) and ((lambda fn, args: [item for res in map( fn, args)
for item in res])(g, phrase)) or (phrase in grammar) and
g(random.choice( grammar[phrase])) or [phrase])('S')
This code defines an anonymous function that takes g
as a parameter and returns another function that takes
phrase as a parameter. The inner function is a modified
version of the original gen function that uses g instead
of gen for the recursive call.
When I ran your code, I got this error:
TypeError: <lambda>() missing 1 required positional
argument: 'g'
I apologize for the error in my previous response. Here's
a corrected version that should work in a single line:
(lambda f: (lambda x: x(x))(lambda y: f(lambda *args: y(y)
( *args))))(lambda g: lambda phrase: isinstance(phrase,
list) and ((lambda fn, args: [item for res in map( fn, args)
for item in res])(g, phrase)) or (phrase in grammar) and
g(random.choice( grammar[phrase])) or [phrase])('S')
This code defines an anonymous function that takes g
as a parameter and returns another function that takes
phrase as a parameter. The inner function is a modified
version of the original gen function that uses g instead
of gen for the recursive call.
You did a good job. I was getting stuck solving it myself.

I'm glad I could help! Don't hesitate to ask if you have

any more questions or if there's anything else I can
assist you with.
 So there you have it: a program with no “=” (so
>>> import mylib
no
>>>variables), but which does a meaningful task!
mylib.displayfile("functional1line.py")
I from
don’trandom
understand this code, but
import choice;from ChatGPT
functools does!
import
reduce
print(reduce(lambda x,y:x+' '+y,(lambda f:
(lambda
x:x(x))(lambda y:f(lambda *a: y(y)(*a))))
(lambda
g:lambda ph: isinstance(ph,list)and((lambda
fn,a:
[item for r in map(fn,a)for item in r])(g,ph))
or(ph in {'S':[['NP','VP']],'NP':[['Art','N']],
'VP':[['V','NP']],'Art':['the','a'],'N':
['table',
'ball','girl','boy'],'V':
>>>
['liked','heard','saw','hit']})
and g(choice({'S':[['NP','VP']],'NP':
>>>
[['Art','N']],
>>>
'VP':[['V','NP']],'Art':['the','a'],'N':
A very complex functional program
% cat functionalExample1line.py
from random import choice;from functools import
reduce
print(reduce(lambda x,y:x+' '+y,(lambda f:(lambda
x:x(x)) (lambda y:f(lambda *a: y(y)(*a))))(lambda
g:lambda ph: isinstance(ph,list)and((lambda fn,a:
[item for r in map(
fn,a)for item in r])(g,ph))or(ph in {'S':[['NP','VP']],
'NP':[['Art','N']],'VP':[['V','NP']],'Art':['the','a'],
'N':['table','ball','girl','boy'],'V':
['liked','heard','saw', 'hit']})and g(choice({'S':
[['NP','VP']],'NP':[['Art','N']], 'VP':
[['V','NP']],'Art':['the','a'],'N':['table','ball',
'girl','boy'],'V':['liked','heard','saw','hit']}[ph]))or
[ph])('S')).capitalize()+".")
% python3 functionalExample1line.py
And this version is true functional programming.
The table saw the ball.
%
Aside from the “import” line, the rest of the code is literally
one long line with no statements. Even I don’t understand it.
The point: Python does support functional programming.
 The built-in functions:
abs() dict() help() min() setattr()
all() dir() hex() next() slice()
any() divmod() id() object() sorted()
ascii() enumerate() input() oct() staticmethod()
bin() eval() int() open() str()
bool() exec() isinstance() ord() sum()
bytearray() filter() issubclass() pow() super()
bytes() float() iter() print() tuple()
callable() format() len() property() type()
chr() frozenset() list() range() vars()
classmethod() getattr() locals() repr() zip()
compile() globals() map() reversed() __import__()
complex() hasattr() max() round()
delattr() hash() memoryview() set()
 The built-in functions:
abs() dict() help() min() setattr()
all() dir() hex() next() slice()
These 11 functions can all be
any() divmod() id() object() sorted()
ascii()
applied to iterables.
enumerate() input() oct() staticmethod()
bin()
That
eval()
makesint()
them well-suited
open()
to str()
bool() use
exec()in functional programming.
isinstance() ord() sum()
bytearray() filter() issubclass() pow() super()
bytes() float() iter() print() tuple()
callable() format() len() property() type()
chr() frozenset() list() range() vars()
classmethod() getattr() locals() repr() zip()
compile() globals() map() reversed() __import__()
complex() hasattr() max() round()
delattr() hash() memoryview() set()
 New topic: An overview of
some important modules
 We’ve just been using functools.reduce
 Earlier today, we used:
math.sin, math.radians, math.factorial,
math.gamma, os.system,
random.randrange, random.choice,
time.time, and time.sleep
 Last week we used:
math.gcd, math.sqrt, math.floor, and
itertools.count
 In even earlier lectures we used:
math.cos, importlib.reload,
 itertools
 infinite(count, cycle, repeat)
 variationsOnBuiltins(slice: islice, zip: zip_longest,
map: starmap, filter: filterfalse)
 makesSmaller(compress, dropwhile, takewhile)
 makesMultiple(groupby, tee)
 combines(combinations, permutations, product, chain)

>>> import itertools; help(itertools)

Help on built-in module itertools:

NAME
itertools - Functional tools for creating and using iterators.

DESCRIPTION
Infinite iterators:
count(start=0, step=1)  start, start+step, start+2*step, ...
cycle(p)  p0, p1, ... plast, p0, p1, ...
repeat(elem [,n])  elem, elem, elem, ... endlessly or up to n times

Iterators terminating on the shortest input sequence:

accumulate(p[, func])  p0, p0+p1, p0+p1+p2
chain(p, q, ...)  p0, p1, ... plast, q0, q1, ...
chain.from_iterable([p, q, ...])  p0, p1, ... plast, q0, q1, ...
:
 compress(data, selectors)  (d[0] if s[0]), (d[1] if s[1]), ...
dropwhile(pred, seq)  seq[n], seq[n+1], starting when pred fails
groupby(iterable[, keyfunc])  sub-iterators grouped by keyfunc(v) value
filterfalse(pred, seq)  elements of seq where pred(elem) is False
islice(seq, [start,] stop [, step])  elements from seq[start:stop:step]
starmap(fun, seq)  fun(*seq[0]), fun(*seq[1]), ...
tee(it, n=2)  (it1, it2 , ... itn) splits one iterator into n
takewhile(pred, seq)  seq[0], seq[1], until pred fails
zip_longest(p, q, ...)  (p[0], q[0]), (p[1], q[1]), ...
Combinatoric generators:
product(p, q, ... [repeat=1])  tuples of all combinations of elements
of p & q & ... As for repeat: product(p,p,p)=product(p,repeat=3)
permutations(p[, r])  tuples of length-r orderings of elements of p
combinations(p, r)  all permutations(p,r) that don't reorder elements
combinations_with_replacement(p, r)  allow elements to be reused
(END)
 itertools
 infinite(count, cycle, repeat)
 variationsOnBuiltins(slice: islice, zip: zip_longest,
map: starmap, filter: filterfalse)
 makesSmaller(compress, dropwhile, takewhile)
 makesMultiple(groupby, tee)
 combines(combinations, permutations, product, chain)

functools
 reduce, partial, partialmethod

>>> import itertools; help (itertools)

>>> from functools import partial#we know reduce
>>> print(partial.__doc__[:-1])
 itertools
 infinite(count, cycle, repeat)
 variationsOnBuiltins(slice: islice, zip: zip_longest,
map: starmap, filter: filterfalse)
 makesSmaller(compress, dropwhile, takewhile)
 makesMultiple(groupby, tee)
 combines(combinations, permutations, product, chain)

functools
 reduce, partial, partialmethod

>>> import itertools; help (itertools)

>>> from functools import partial#we know reduce
>>> print(partial.__doc__[:-1])
partial(func, *args, **keywords) – return
 itertools
 infinite(count, cycle, repeat)
 variationsOnBuiltins(slice: islice, zip: zip_longest,
map: starmap, filter: filterfalse)
 makesSmaller(compress, dropwhile, takewhile)
 makesMultiple(groupby, tee)
 combines(combinations, permutations, product, chain)

functools
 reduce, partial, partialmethod

>>> import itertools; help (itertools)

>>> from functools import partial#we know reduce
>>> print(partial.__doc__[:-1])
partial(func, *args, **keywords) – return
a new function with partial application
 itertools
 infinite(count, cycle, repeat)
 variationsOnBuiltins(slice: islice, zip: zip_longest,
map: starmap, filter: filterfalse)
 makesSmaller(compress, dropwhile, takewhile)
 makesMultiple(groupby, tee)
 combines(combinations, permutations, product, chain)

functools
 reduce, partial, partialmethod

>>> import itertools; help (itertools)

>>> from functools import partial#we know reduce
>>> print(partial.__doc__[:-1])
partial(func, *args, **keywords) – return
a new function with partial application
of the given arguments and keywords.
 itertools
 infinite(count, cycle, repeat)
 variationsOnBuiltins(slice: islice, zip: zip_longest,
map: starmap, filter: filterfalse)
 makesSmaller(compress, dropwhile, takewhile)
 makesMultiple(groupby, tee)
 combines(combinations, permutations, product, chain)

functools
 reduce, partial, partialmethod

>>> from functools import partial#we know reduce

>>> print(partial.__doc__[:-1])
partial(func, *args, **keywords) – return
a new function with partial application
of the given arguments and keywords.
>>> indpnt=partial(print,'[\t',']')
 itertools
 infinite(count, cycle, repeat)
 variationsOnBuiltins(slice: islice, zip: zip_longest,
map: starmap, filter: filterfalse)
 makesSmaller(compress, dropwhile, takewhile)
 makesMultiple(groupby, tee)
 combines(combinations, permutations, product, chain)

functools
 reduce, partial, partialmethod

>>> print(partial.__doc__[:-1])
partial(func, *args, **keywords) – return
a new function with partial application
of the given arguments and keywords.
>>> indpnt=partial(print,'[\t',']')
>>> indpnt("Hello world!")
 itertools
 infinite(count, cycle, repeat)
 variationsOnBuiltins(slice: islice, zip: zip_longest,
map: starmap, filter: filterfalse)
 makesSmaller(compress, dropwhile, takewhile)
 makesMultiple(groupby, tee)
 combines(combinations, permutations, product, chain)

functools
 reduce, partial, partialmethod

partial(func, *args, **keywords) – return

a new function with partial application
of the given arguments and keywords.
>>> indpnt=partial(print,'[\t',']')
>>> indpnt("Hello world!")
[ ] Hello world!
 itertools
 infinite(count, cycle, repeat)
 variationsOnBuiltins(slice: islice, zip: zip_longest,
map: starmap, filter: filterfalse)
 makesSmaller(compress, dropwhile, takewhile)
 makesMultiple(groupby, tee)
I won’t test you on this,
 combines(combinations, permutations, product, chain)
but this how to get red
functools text in UNIX (my Python
 reduce, partial, partialmethod is running in UNIX).
a new function with partial application
of the given arguments and keywords.
>>> indpnt=partial(print,'[\t',']')
>>> indpnt("Hello world!")
[ ] Hello world!
>>> copnt=partial(print,'\033[7;41;30m→',sep=
 itertools
 infinite(count, cycle, repeat)
 variationsOnBuiltins(slice: islice, zip: zip_longest,
map: starmap, filter: filterfalse)
 makesSmaller(compress, dropwhile, takewhile)
 makesMultiple(groupby, tee)
 combines(combinations, permutations, product, chain)
I won’t test you on this,
functools but this how to get green
 reduce, partial, partialmethod italics text in UNIX.
of the given arguments and keywords.
>>> indpnt=partial(print,'[\t',']')
>>> indpnt("Hello world!")
[ ] Hello world!
>>> copnt=partial(print,'\033[7;41;30m→',sep=
... '\033[3;30;42m ',end='\033[1;m\n')
 itertools
 infinite(count, cycle, repeat)
 variationsOnBuiltins(slice: islice, zip: zip_longest,
map: starmap, filter: filterfalse)
 makesSmaller(compress, dropwhile, takewhile)
 makesMultiple(groupby, tee)
 combines(combinations, permutations, product, chain)

functools
 reduce, partial, partialmethod

>>> indpnt=partial(print,'[\t',']')
>>> indpnt("Hello world!")
[ ] Hello world!
>>> copnt=partial(print,'\033[7;41;30m→',sep=
... '\033[3;30;42m ',end='\033[1;m\n')
>>> copnt('Hello','World',1,2,3,[*"abc"])
 itertools
 infinite(count, cycle, repeat)
 variationsOnBuiltins(slice: islice, zip: zip_longest,
map: starmap, filter: filterfalse)
 makesSmaller(compress, dropwhile, takewhile)
 makesMultiple(groupby, tee)
 combines(combinations, permutations, product, chain)

functools
 reduce, partial, partialmethod

>>> indpnt("Hello world!")

[ ] Hello world!
>>> copnt=partial(print,'\033[7;41;30m→',sep=
... '\033[3;30;42m ',end='\033[1;m\n')
>>> copnt('Hello','World',1,2,3,[*"abc"])
→ Hello World 1 2 3 ['a', 'b', 'c']
 itertools
 infinite(count, cycle, repeat)
 variationsOnBuiltins(slice: islice, zip: zip_longest,
map: starmap, filter: filterfalse)
 makesSmaller(compress, dropwhile, takewhile)
 makesMultiple(groupby, tee)
 combines(combinations, permutations, product, chain)

functools
 reduce, partial, partialmethod

[ ] Hello world!
>>> copnt=partial(print,'\033[7;41;30m→',sep=
... '\033[3;30;42m ',end='\033[1;m\n')
>>> copnt('Hello','World',1,2,3,[*"abc"])
→ Hello World 1 2 3 ['a', 'b', 'c']
>>>
itertools
 infinite(count, cycle, repeat)
 variationsOnBuiltins(slice: islice, zip: zip_longest,
map: starmap, filter: filterfalse)
 makesSmaller(compress, dropwhile, takewhile)
 makesMultiple(groupby, tee)
 combines(combinations, permutations, product, chain)

functools
 reduce, partial, partialmethod
itertools
 infinite(count, cycle, repeat)
 variationsOnBuiltins(slice: islice, zip: zip_longest,
map: starmap, filter: filterfalse)
 makesSmaller(compress, dropwhile, takewhile)
 makesMultiple(groupby, tee)
 combines(combinations, permutations, product, chain)

functools
 reduce, partial, partialmethod

operator
 ComparisonOps (lt, eq, le, ne, gt, ge)
 LogicOps(not_, or_, and_, xor)
 UnaryOps(abs, inv, neg, truth)
 BinaryOps(concat, add, sub, mul, mod, truediv, pow)
 ContainerOps(contains, index, indexOf, countOf)
itertools
functools.reduce(function, sequence)

Sl
id
 infinite(count, cycle, repeat)
– Apply the function to each item in the sequence.

e
25
 variationsOnBuiltins(slice:
– islice, zip: zip_longest,
The function must take 2 arguments & return 1 value.
– That returned valuemap: starmap,
becomes the 1st filter: filterfalse)
argument to the function
on the next iteration (2nd argument
 makesSmaller(compress, is thetakewhile)
dropwhile, next sequence item).
– Return 1 value, the finaltee)
 makesMultiple(groupby, value returned by the function.
 >>> from functools import reduce
combines(combinations, permutations, product, chain)
>>> reduce(lambda x,y: x+y, range(9))
36
functools
>>> reduce(lambda x,y: x+y, map(str,range(9))))
 reduce, partial, partialmethod
'012345678'
>>> print(reduce.__doc__[:46]+reduce.__doc__[47:])
operator
reduce(function, sequence [, initial]) -> value
 ComparisonOps (lt, eq, le, ne, gt, ge)
Apply a function of 2 arguments cumulatively to the
 items
LogicOps(not_, or_, and_,
of a sequence, xor) to right, so as to reduce
from left
 the sequence to inv,
UnaryOps(abs, a single value. For example,
neg, truth)
reduce(lambda x, y: x+y, [1, 2, 3, 4, 5]) calculates
 BinaryOps(concat, add, sub, mul, mod, truediv, pow)
((((1+2)+3)+4)+5). If initial is present, it is placed
 before
ContainerOps(contains, index, indexOf,
the items of the sequence in the countOf)
calculation, and
>>> reduce(lambda x,y: x+y, range(9))
itertools
 functools.reduce(function, sequence)
36
 infinite(count, cycle, repeat)
>>> –reduce(lambda
Apply the function x,y: x+y, item
to each map(str,range(9))))
in the sequence.
'012345678'
 variationsOnBuiltins(slice: islice, zip:&zip_longest,
– The function must take 2 arguments return 1 value.
>>> –
That returned valuemap: starmap,
becomes the 1st filter: filterfalse)
argument to the function
on the next iteration (2nd argument
 makesSmaller(compress, is thetakewhile)
dropwhile, next sequence item).
–Return 1 value, the finaltee)
 makesMultiple(groupby, value returned by the function.
 >>> from functools import reduce
combines(combinations, permutations, product, chain)
>>> reduce(lambda x,y: x+y, range(9))
36
functools
>>> reduce(lambda x,y: x+y, map(str,range(9))))
 reduce, partial, partialmethod
'012345678'
>>> print(reduce.__doc__[:46]+reduce.__doc__[47:])
operator
reduce(function, sequence [, initial]) -> value
 ComparisonOps (lt, eq, le, ne, gt, ge)
Apply a function of 2 arguments cumulatively to the
 items
LogicOps(not_, or_, and_,
of a sequence, xor) to right, so as to reduce
from left
 the sequence to inv,
UnaryOps(abs, a single value. For example,
neg, truth)
reduce(lambda x, y: x+y, [1, 2, 3, 4, 5]) calculates
 BinaryOps(concat, add, sub, mul, mod, truediv, pow)
((((1+2)+3)+4)+5). If initial is present, it is placed
 before
ContainerOps(contains, index, indexOf,
the items of the sequence in the countOf)
calculation, and
>>> reduce(lambda x,y: x+y, range(9))
itertools
36
 infinite(count, cycle, repeat)
>>> reduce(lambda x,y: x+y, map(str,range(9))))
'012345678'
 variationsOnBuiltins(slice: islice, zip: zip_longest,
>>> sum(range(9)) map: starmap, filter: filterfalse)
 makesSmaller(compress, dropwhile, takewhile)
 makesMultiple(groupby, tee)
 combines(combinations, permutations, product, chain)

functools
 reduce, partial, partialmethod

operator
 ComparisonOps (lt, eq, le, ne, gt, ge)
 LogicOps(not_, or_, and_, xor)
 UnaryOps(abs, inv, neg, truth)
 BinaryOps(concat, add, sub, mul, mod, truediv, pow)
 ContainerOps(contains, index, indexOf, countOf)
>>> reduce(lambda x,y: x+y, range(9))
itertools
36
 infinite(count, cycle, repeat)
>>> reduce(lambda x,y: x+y, map(str,range(9))))
'012345678'
 variationsOnBuiltins(slice: islice, zip: zip_longest,
>>> sum(range(9)) map: starmap, filter: filterfalse)
36
 makesSmaller(compress,
>>> try: sum("a","b") dropwhile, takewhile)
... except: print("Can't tee)
 makesMultiple(groupby, use sum() on strings")
...
 combines(combinations,
Can't use sum() on strings permutations, product, chain)
>>> import operator as op
functools
>>>
 reduce, partial, partialmethod
reduce(op.add,map(str,range(9))),reduce(op.add,range(9
))
>>>
operator
('012345678', 36)
>>> reduce(op.concat,
 ComparisonOps map(str,range(9)))
(lt, eq, le, ne, gt, ge)
'012345678'
...
 LogicOps(not_, or_, and_, xor)
>>>
... try: reduce(op.concat, range(9))
 UnaryOps(abs,
... except: print("Can't use concat() on numbers")
inv, neg, truth)
>>>
 BinaryOps(concat,
concat on add
add, sub, mul, mod, truediv, pow)
Can't use concat() numbers
 ContainerOps(contains,
>>> "".join(map(str,range(9))) index, indexOf, countOf)
#BTW, this also worx
collections # These extend the 7(or 9) built-in data types
 namedtuple, deque, defaultdict, OrderedDict, Counter
>>> print(collections.__doc__)
This module implements specialized container datatypes providing alternatives
to Python's general purpose built-in containers, dict, list, set, and tuple.

* namedtuple factory function for creating tuple subclasses with named fields
* deque list-like container with fast appends and pops on either end
* ChainMap dict-like class for creating a single view of multiple mappings
* Counter dict subclass for counting hashable objects
* OrderedDict dict subclass that remembers the order entries were added
* defaultdict dict subclass that calls a factory function to get missing values
* UserDict wrapper around dictionary objects for easier dict subclassing
* UserList wrapper around list objects for easier list subclassing
* UserString wrapper around string objects for easier string subclassing
>>>
https://docs.python.org/3/library/collections.html
collections # These extend the 7(or 9) built-in data types
 namedtuple, deque, defaultdict, OrderedDict, Counter

bisect # Optimizes handling of sorted lists

 bisect, insort
>>> import bisect as bs
>>> print(bs.bisect.__doc__[:294]+".\n")
bisect_right(a, x[, lo[, hi]]) -> index
Return the index where to insert item x in list a, assuming a is sorted.
The return value i is such that all e in a[:i] have e <= x, and all e in a[i:] have e >
x. So if x is already in the list, i points just beyond the rightmost x already
there.
>>>
>>> print(bs.insort.__doc__[:163])
insort_right(a, x[, lo[, hi]])
Insert item x in list a, and keep it sorted assuming a is sorted.
If x is already in a, insert it to the right of the rightmost x.
collections # These extend the 7(or 9) built-in data types
 namedtuple, deque, defaultdict, OrderedDict, Counter

bisect # Optimizes handling of sorted lists

 bisect, insort

copy #shallow vs deep-copy = Do contained items get

duplicated?
 copy, deepcopy
>>> import copy as cp
>>> print(cp.__doc__[48:67]+cp.__doc__[89:204]+cp.__doc__[255:698])
Interface summary:
x = copy.copy(y) # make a shallow copy of y
x = copy.deepcopy(y) # make a deep copy of y
The difference between shallow and deep copying is only relevant for compound
objects (objects that contain other objects, like lists or class instances).
- A shallow copy constructs a new compound object and then (to the extent
possible) inserts *the same objects* into it that the original contains.
collections # These extend the 7(or 9) built-in data types
 namedtuple, deque, defaultdict, OrderedDict, Counter

bisect # Optimizes handling of sorted lists

 bisect, insort

copy #shallow vs deep-copy = Do contained items get

duplicated?
 copy, deepcopy

math / cmath (for complex numbers) # All the math funcs

random # As we've used in the homework
 random, uniform, randint, randrange, choice

importlib # See lecture 4, slide 107

 reload

keyword #Just lists Python's keywords. (We'll use it today…)

 kwlist, iskeyword