TLE in FERT21_0 (Matches)

[deep_shredder]

this is my code

import sys,math
inp=sys.stdin.read().split()
for x in range(int(inp[0])):
   n=int(inp[x+1])
   if n==1:print ('1');continue
   x=5**(n-1)
   l=int(math.log10(x))+1
   op="0.%s%d" % ('0'*(n-1-l),x)
   print(op)

my code runs till the 10th case and then times out.
is the problem io related or do i have to completely change my approach?

[numerix]

Your code is about 10 times slower than my AC code with the worst input data.
Implementation should be optimized.
According to I/O it's not a matter of large I/O but of expensive type conversions.

[deep_shredder]

now i am finding the powers using bit operators

import sys,math
inp=input()
inp=sys.stdin.read().split()
for n in inp:
   n=int(n)-1
   if n==0:print ('1');continue
   x=2<<(n-1)
   f="%."+str(n)+"f"
   op=f% (1.0/x)
   print(op)

I get
TLE -python 3.1.2
NZEC - python 2.7

i will try and reduce type conversions and get back

[deep_shredder]

import sys,math
inp=input()
inp=sys.stdin.read().split()
for n in inp:
   n=int(n)-1
   if n==0:print ('1');continue
   x=1<<n
   f= "%%.%df" % (n)
   op=f% (1.0/x)
   print(op)

i think i have reduced the type conversions. Help me find the bottleneck.

[daft_wullie]

What result do you get for your 2.7-Code? really TLE?

I'm asking because of this.

IDEONE uses Python 2.7. With Python 2.5/2.6 you get the following:

[code]>>> "%.100f" %(1.0/(1<Traceback (most recent call last):
File "", line 1, in
OverflowError: formatted float is too long (precision too large?)
[/code]

So your code might not be able to run, unless it is assigned to a python 2.7 machine.

And I noticed that all my recent python submissions have been processed by python 2.5 servers.

I did not resolve this problem up to now - my only ACC code was run on a 2.7 machine and all re-submissions resulted in NZEC.