Recursion Question in Python. No Conditionals or Loops
Question:
I am trying to figure out how to print the word "hello" 121 times in python. I need to use a function without conditionals or loops, no new lines, and no multiplying the string by an integer.
I thinking something like:
print_hello():
print('hello')
print_hello()
print_hello()
but I can’t seem to find a way to limit the recursive output without conditionals. Any help would be greatly appreciated.
Update
Here are all of the constraints to the challenge, it’s possible recursion isn’t the right approach.
- max 20 lines of code
- no inputs to a function.
- no imports or modules
- no if statements
- no loops (for or while)
- no new line characters (n)
- can’t do something like: print("hellon"*121)
- no semi colons or tuples
- can’t use exec functions
Answers:
from sys import setrecursionlimit
setrecursionlimit(121)
def phello():
print('hello', end='')
phello()
try:
phello()
except:
print('done')
def print_hello():
print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello');
def print_hello(i):
print('hello', end="")
i -= 1
a = 42 / i
try:
print_hello(i)
except ZeroDivisionError:
pass
print_hello(121)
The 42 could be anything really, just need to trigger the division by 0 error
OK. Just another answer, just for fun.
def print_hello121():
print(('n'.join(['hello']*121)))
Here – we multiply the list, not the string.
def print_single_hello():
print('hello')
def print_hello121():
[print_single_hello()] * 121
This does not actually works correctly, but it can be an idea.
Just using lists its pretty easy:
def myr(l=[lambda *a: 0] + 120*[lambda a, b: a(b)]):
print("hello", end="")
l.pop()(myr, l)
myr()
The idea.
def loop(x):
try:
next(x)
print("hello", end="")
loop(x)
except StopaIteration:
pass
def hello(n_repeat):
loop(iter(range(n_repeat)))
...
hello(50)
Using lazy operators and
or or
, it is easy to fake a conditional in python:
def print_hello(n):
n > 0 and (print('hello') or print_hello(n - 1))
print_hello(3)
# hello
# hello
# hello
print_hello(121)
# ...
You can use the short circuiting property of and
. Zero is False, every other number is True. If the first part of the expression is false the second part isn’t evaluated.
def print_hello(n=121):
print('hello', end=' ')
n and print_hello(n - 1)
print_hello()
Edit: I saw that you disqualified another answer because the function had a parameter, so here’s a version that uses a global variable instead. It also fixes an off-by-one problem in the original.
n = 121
def print_hello():
global n
print('hello', end=' ')
n -= 1
n and print_hello()
print_hello()
(Note: I’ve edited the message. So see better solution at the end)
If we don’t try to find a silver bullet, using a trick that would have been forgotten in the restrictions (global
variables to bypass parameters interdiction, and
/try
to bypass if
interdiction, multiplying list to bypass multiplication of string interdiction, …), as I understand the question, it is about finding a correct splitting in functions/subfunctions to have 20 lines of code (no semicolon)
For example, if we had to print 32 hello, we could, as geeks used to reason in power of 2,
def h2():
print("hello")
print("hello")
def h4():
h2()
h2()
def h8():
h4()
h4()
def h16():
h8()
h8()
h16()
h16()
Which are 14 lines.
But the def
parts complicates things, and complicates what is optimal. For example, here, since we don’t use h2
elsewhere that in h4
, it would be shorter to directly print hello
4 times in h4
. Likewise, for h16
def h4():
print("hello")
print("hello")
print("hello")
print("hello")
def h16():
h4()
h4()
h4()
h4()
h16()
h16()
which are only 12 lines.
Now, here, number is 121. That is a specific number. Which is even the reason why I believe that this is the kind of expected solution: it is not easy to decide what intermediate function we need to create.
That would be an interesting problem per se: create a code, that optimize the number of lines needed, using this kind of subfunctions encapsulation.
But one combination that fit in 20 lines is
def h3():
print("hello")
print("hello")
print("hello")
def h7():
h3()
h3()
print("hello")
def h15():
h7()
h7()
print("hello")
def h60():
h15()
h15()
h15()
h15()
h60()
h60()
print("hello")
I know it is way less smart (and, even, "smart ass") than all the other solutions that were proposed. But I am really convinced that this is the kind of expected solution. It may seem too simple and naive. But it is not that an easy problem to decide which h??
to write (well, it is not that hard if the constraint is just "fit in 20 lines". But I would have a harder time if the constraint was "use the smallest number of lines possible")
Edit
I couldn’t resist, so I wrote a code that optimize this
def size(n, l):
ml=max(k for k in l if k<=n)
nm=n//ml
r=n-nm*ml
if r==0:
return nm
else:
return nm+size(r, l)
def sizefn(l):
return sum(size(k, [x for x in l if x<k])+1 for k in l if k>1)
def totsize(n,l):
return size(n, l) + sizefn(l)
rec=1000
def compute(l, k):
global rec
if k>120: return False
sfn =sizefn(l)
if sfn+2>=rec:
return False
f = size(121, l)
if sfn+f<rec:
rec=sfn+f
print(f'{sfn+f} ({sfn}+{f}) :', l)
compute(l+[k], k+1)
compute(l, k+1)
What it does is just try all possible combinations of intermediate functions. So, that is, theoretically 2¹²⁰ combinations (all intermediate function h? may exist or not), and count how many line of code that would be, and keep the best.
Except that I do it with Branch&Bound, allowing to avoid examination of whole subsets of the set of all combination.
Result is
121 (0+121) : [1]
64 (3+61) : [1, 2]
47 (6+41) : [1, 2, 3]
40 (9+31) : [1, 2, 3, 4]
37 (12+25) : [1, 2, 3, 4, 5]
36 (15+21) : [1, 2, 3, 4, 5, 6]
35 (31+4) : [1, 2, 3, 4, 5, 6, 7, 8, 9, 13, 39]
34 (28+6) : [1, 2, 3, 4, 5, 6, 7, 8, 9, 23]
33 (28+5) : [1, 2, 3, 4, 5, 6, 7, 8, 10, 30]
32 (28+4) : [1, 2, 3, 4, 5, 6, 7, 8, 13, 39]
31 (25+6) : [1, 2, 3, 4, 5, 6, 7, 8, 23]
30 (25+5) : [1, 2, 3, 4, 5, 6, 7, 10, 30]
29 (25+4) : [1, 2, 3, 4, 5, 6, 7, 13, 39]
28 (22+6) : [1, 2, 3, 4, 5, 6, 8, 24]
27 (22+5) : [1, 2, 3, 4, 5, 6, 10, 30]
26 (20+6) : [1, 2, 3, 4, 5, 6, 23]
25 (19+6) : [1, 2, 3, 4, 5, 8, 24]
24 (19+5) : [1, 2, 3, 4, 5, 10, 30]
23 (17+6) : [1, 2, 3, 4, 6, 24]
22 (16+6) : [1, 2, 3, 4, 8, 24]
21 (16+5) : [1, 2, 3, 5, 10, 30]
20 (14+6) : [1, 2, 3, 6, 24]
19 (13+6) : [1, 2, 4, 8, 24]
18 (12+6) : [1, 2, 6, 24]
And it ends there.
Meaning that there is no better solution that the one with h2, h6, and h24 (h1 is just print(hello)
)
For a total of 18 lines
Which gives
def h2():
print("hello")
print("hello")
def h6():
h2()
h2()
h2()
def h24():
h6()
h6()
h6()
h6()
h24()
h24()
h24()
h24()
h24()
print("hello")
You could use a "higher-order function" such as map
, filter
, max
, min
or sorted
, to apply a function repeatedly to a range
.
def print_hello_map(n):
any(map(lambda _: print('hello'), range(n)))
def print_hello_filter(n):
next(filter(lambda x: print('hello') or x + 1>= n, range(n)))
def print_hello_max(n):
max(range(n), key=lambda x: bool(print('hello')))
def print_hello_sorted(n):
sorted(range(n), key=lambda x: bool(print('hello')))
If you had allowed imports from the standard library, then there would have been many alternatives to map
and filter
, for instance functools.reduce
and itertools.repeat
, but also all functions from itertools
that perform any kind of looping.
from functools import reduce
def print_hello_reduce(n):
reduce(lambda acc,x: print('hello'), range(n+1))
from itertools import repeat
def f(x):
yield bool(print('hello'))
def print_hello_repeat(n):
sum(repeat(f, n))
Try this, you are thinking too hard:
x='hello'
print(x + ' 121 times')
I am trying to figure out how to print the word "hello" 121 times in python. I need to use a function without conditionals or loops, no new lines, and no multiplying the string by an integer.
I thinking something like:
print_hello():
print('hello')
print_hello()
print_hello()
but I can’t seem to find a way to limit the recursive output without conditionals. Any help would be greatly appreciated.
Update
Here are all of the constraints to the challenge, it’s possible recursion isn’t the right approach.
- max 20 lines of code
- no inputs to a function.
- no imports or modules
- no if statements
- no loops (for or while)
- no new line characters (n)
- can’t do something like: print("hellon"*121)
- no semi colons or tuples
- can’t use exec functions
from sys import setrecursionlimit
setrecursionlimit(121)
def phello():
print('hello', end='')
phello()
try:
phello()
except:
print('done')
def print_hello():
print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello'); print('hello');
def print_hello(i):
print('hello', end="")
i -= 1
a = 42 / i
try:
print_hello(i)
except ZeroDivisionError:
pass
print_hello(121)
The 42 could be anything really, just need to trigger the division by 0 error
OK. Just another answer, just for fun.
def print_hello121():
print(('n'.join(['hello']*121)))
Here – we multiply the list, not the string.
def print_single_hello():
print('hello')
def print_hello121():
[print_single_hello()] * 121
This does not actually works correctly, but it can be an idea.
Just using lists its pretty easy:
def myr(l=[lambda *a: 0] + 120*[lambda a, b: a(b)]):
print("hello", end="")
l.pop()(myr, l)
myr()
The idea.
def loop(x):
try:
next(x)
print("hello", end="")
loop(x)
except StopaIteration:
pass
def hello(n_repeat):
loop(iter(range(n_repeat)))
...
hello(50)
Using lazy operators and
or or
, it is easy to fake a conditional in python:
def print_hello(n):
n > 0 and (print('hello') or print_hello(n - 1))
print_hello(3)
# hello
# hello
# hello
print_hello(121)
# ...
You can use the short circuiting property of and
. Zero is False, every other number is True. If the first part of the expression is false the second part isn’t evaluated.
def print_hello(n=121):
print('hello', end=' ')
n and print_hello(n - 1)
print_hello()
Edit: I saw that you disqualified another answer because the function had a parameter, so here’s a version that uses a global variable instead. It also fixes an off-by-one problem in the original.
n = 121
def print_hello():
global n
print('hello', end=' ')
n -= 1
n and print_hello()
print_hello()
(Note: I’ve edited the message. So see better solution at the end)
If we don’t try to find a silver bullet, using a trick that would have been forgotten in the restrictions (global
variables to bypass parameters interdiction, and
/try
to bypass if
interdiction, multiplying list to bypass multiplication of string interdiction, …), as I understand the question, it is about finding a correct splitting in functions/subfunctions to have 20 lines of code (no semicolon)
For example, if we had to print 32 hello, we could, as geeks used to reason in power of 2,
def h2():
print("hello")
print("hello")
def h4():
h2()
h2()
def h8():
h4()
h4()
def h16():
h8()
h8()
h16()
h16()
Which are 14 lines.
But the def
parts complicates things, and complicates what is optimal. For example, here, since we don’t use h2
elsewhere that in h4
, it would be shorter to directly print hello
4 times in h4
. Likewise, for h16
def h4():
print("hello")
print("hello")
print("hello")
print("hello")
def h16():
h4()
h4()
h4()
h4()
h16()
h16()
which are only 12 lines.
Now, here, number is 121. That is a specific number. Which is even the reason why I believe that this is the kind of expected solution: it is not easy to decide what intermediate function we need to create.
That would be an interesting problem per se: create a code, that optimize the number of lines needed, using this kind of subfunctions encapsulation.
But one combination that fit in 20 lines is
def h3():
print("hello")
print("hello")
print("hello")
def h7():
h3()
h3()
print("hello")
def h15():
h7()
h7()
print("hello")
def h60():
h15()
h15()
h15()
h15()
h60()
h60()
print("hello")
I know it is way less smart (and, even, "smart ass") than all the other solutions that were proposed. But I am really convinced that this is the kind of expected solution. It may seem too simple and naive. But it is not that an easy problem to decide which h??
to write (well, it is not that hard if the constraint is just "fit in 20 lines". But I would have a harder time if the constraint was "use the smallest number of lines possible")
Edit
I couldn’t resist, so I wrote a code that optimize this
def size(n, l):
ml=max(k for k in l if k<=n)
nm=n//ml
r=n-nm*ml
if r==0:
return nm
else:
return nm+size(r, l)
def sizefn(l):
return sum(size(k, [x for x in l if x<k])+1 for k in l if k>1)
def totsize(n,l):
return size(n, l) + sizefn(l)
rec=1000
def compute(l, k):
global rec
if k>120: return False
sfn =sizefn(l)
if sfn+2>=rec:
return False
f = size(121, l)
if sfn+f<rec:
rec=sfn+f
print(f'{sfn+f} ({sfn}+{f}) :', l)
compute(l+[k], k+1)
compute(l, k+1)
What it does is just try all possible combinations of intermediate functions. So, that is, theoretically 2¹²⁰ combinations (all intermediate function h? may exist or not), and count how many line of code that would be, and keep the best.
Except that I do it with Branch&Bound, allowing to avoid examination of whole subsets of the set of all combination.
Result is
121 (0+121) : [1]
64 (3+61) : [1, 2]
47 (6+41) : [1, 2, 3]
40 (9+31) : [1, 2, 3, 4]
37 (12+25) : [1, 2, 3, 4, 5]
36 (15+21) : [1, 2, 3, 4, 5, 6]
35 (31+4) : [1, 2, 3, 4, 5, 6, 7, 8, 9, 13, 39]
34 (28+6) : [1, 2, 3, 4, 5, 6, 7, 8, 9, 23]
33 (28+5) : [1, 2, 3, 4, 5, 6, 7, 8, 10, 30]
32 (28+4) : [1, 2, 3, 4, 5, 6, 7, 8, 13, 39]
31 (25+6) : [1, 2, 3, 4, 5, 6, 7, 8, 23]
30 (25+5) : [1, 2, 3, 4, 5, 6, 7, 10, 30]
29 (25+4) : [1, 2, 3, 4, 5, 6, 7, 13, 39]
28 (22+6) : [1, 2, 3, 4, 5, 6, 8, 24]
27 (22+5) : [1, 2, 3, 4, 5, 6, 10, 30]
26 (20+6) : [1, 2, 3, 4, 5, 6, 23]
25 (19+6) : [1, 2, 3, 4, 5, 8, 24]
24 (19+5) : [1, 2, 3, 4, 5, 10, 30]
23 (17+6) : [1, 2, 3, 4, 6, 24]
22 (16+6) : [1, 2, 3, 4, 8, 24]
21 (16+5) : [1, 2, 3, 5, 10, 30]
20 (14+6) : [1, 2, 3, 6, 24]
19 (13+6) : [1, 2, 4, 8, 24]
18 (12+6) : [1, 2, 6, 24]
And it ends there.
Meaning that there is no better solution that the one with h2, h6, and h24 (h1 is just print(hello)
)
For a total of 18 lines
Which gives
def h2():
print("hello")
print("hello")
def h6():
h2()
h2()
h2()
def h24():
h6()
h6()
h6()
h6()
h24()
h24()
h24()
h24()
h24()
print("hello")
You could use a "higher-order function" such as map
, filter
, max
, min
or sorted
, to apply a function repeatedly to a range
.
def print_hello_map(n):
any(map(lambda _: print('hello'), range(n)))
def print_hello_filter(n):
next(filter(lambda x: print('hello') or x + 1>= n, range(n)))
def print_hello_max(n):
max(range(n), key=lambda x: bool(print('hello')))
def print_hello_sorted(n):
sorted(range(n), key=lambda x: bool(print('hello')))
If you had allowed imports from the standard library, then there would have been many alternatives to map
and filter
, for instance functools.reduce
and itertools.repeat
, but also all functions from itertools
that perform any kind of looping.
from functools import reduce
def print_hello_reduce(n):
reduce(lambda acc,x: print('hello'), range(n+1))
from itertools import repeat
def f(x):
yield bool(print('hello'))
def print_hello_repeat(n):
sum(repeat(f, n))
Try this, you are thinking too hard:
x='hello'
print(x + ' 121 times')