Member Avatar for nunos

Hi I would like to know if there is any way to obtain a float result from a division between two integers other than using "from __future__ import division"?

I am not looking for the .0 'trick'. Example 4/5.0 = 0.8.

I want to be able to do 4/5 = 0.8.

Thanks.

  • 4 Contributors
  • 7 Replies
  • 202 Views
  • 22 Hours Discussion Span
  • Latest Post Latest Post by nunos

Dani AI

Generated

Short answer: you can't "unload" the future-division behavior once code is compiled with it — the future statement is a compile-time directive that applies to the whole module (or to any code objects compiled with the same future flags). If you need both classic (integer) and true (floating) division in one program, choose one of these patterns rather than trying to toggle the future import at runtime. (larsen-family.us)

What the replies already showed is correct in spirit. The simplest and most maintainable approaches are explicit: either make one operand a floating type when you want a float (this is the quick inline fix many suggested), or move code that needs true-division into a separate module and give that module the new semantics. The language change that made true-division the default in Python 3 (and introduced an unambiguous floor operator) is documented in PEP 238 — it's why mixing semantics across a codebase is risky. (peps.python.org)

If you must evaluate a small expression with "Python‑3 style" division without changing the surrounding module, compile and run that expression with the appropriate future flag. For example:

import __future__
expr = "num / den"
code = compile(expr, "<expr>", "eval", flags=__future__.division.compiler_flag)
print(eval(code, {"num": 4, "den": 5}))   # evaluates with true-division semantics

This uses the feature bit exposed by the future module and the compile() flags to control semantics for the compiled code object. Use this sparingly — it adds complexity but lets you localize behavior. (docs.python.org)

Recommendation: for clarity and future-proofing prefer explicit casts or run the code under Python 3 (where / is true division and // is floor division). If you must mix both semantics, split the code into separate modules (one with the new semantics, one without) or use targeted compile() calls as shown above. Avoid string-manipulation hacks for decimal output — they’re brittle, especially for negative values and edge cases. (peps.python.org)

Recommended Answers

All 7 Replies

Member Avatar for wildgoose

No matter how you do it, the fraction part is the result of the remainder / divisor so still need a division!

Could fake it somewhat.
In integer
part %
----- = -------
whole 100

So could calculate a whole integer form of the remainder and stick your own decimal point as a string character!

(4 * 100) / 5 = 80

4 / 5 = 0
"0" "." "80" 0.80

Member Avatar for nunos

No matter how you do it, the fraction part is the result of the remainder / divisor so still need a division!

Could fake it somewhat.
In integer
part %
----- = -------
whole 100

So could calculate a whole integer form of the remainder and stick your own decimal point as a string character!

(4 * 100) / 5 = 80

4 / 5 = 0
"0" "." "80" 0.80

I was just wondering if there was any other way other than "from __future__ import division" because further in the program I might want to have 4/5 = 0, and because of that I won't.

Is it possible to unload from __future__ import division?

Member Avatar for wildgoose

No idea. Sorry!

Member Avatar for nunos

No idea. Sorry!

I guess I will have to do the .0 'trick'. Maybe some other folk will know a way.

Thanks wildgoose for your help.

Member Avatar for jlm699

I guess I will have to do the .0 'trick'. Maybe some other folk will know a way.

Thanks wildgoose for your help.

You could do this:

float(4)/5

Or if you wanted to perform the from __future__ import division and still get integer returns you could simply turn the answer into an int like this:

>>> ans = 5 / 2
>>> ans
2.5
>>> int(ans)
2
>>>
Member Avatar for vegaseat

You could update to Python 3.1 where '/' is the floatingpoint division and '//' is the integer division.

Actually, with Python2 versions, 4./5 would be simpler to type then 4.0/5 or float(4)/5 :)

Member Avatar for nunos

You could update to Python 3.1 where '/' is the floatingpoint division and '//' is the integer division.

Actually, with Python2 versions, 4./5 would be simpler to type then 4.0/5 or float(4)/5 :)

Thanks for the two tips. I will keep that in mind.

Cheers :icon_wink:

Be a part of the DaniWeb community

We're a friendly, industry-focused community of developers, IT pros, digital marketers, and technology enthusiasts meeting, networking, learning, and sharing knowledge.