Last decimal digit precision changes in different call of same generator function [python] [duplicate]

Question

I created this generator function:

def myRange(start,stop,step):
    r = start
    while r < stop:
        yield r
        r += step

and I use it in two different ways. 1st:

for x in myRange(0,1,0.1):
    print x

Result:

0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0

2nd way to call the function:

a = [x for x in myRange(0,1,0.1)]

which results in:

[0, 0.1, 0.2, 0.30000000000000004, 0.4, 0.5, 0.6, 0.7, 0.7999999999999999, 0.8999999999999999, 0.9999999999999999]

Why are the values generated different?

Martijn Pieters · Accepted Answer · 2014-10-21 17:00:23Z

8

It is not the order in which you called your generator, but the way you are presenting the numbers that caused this change in output.

You are printing a list object the second time, and that's a container. Container contents are printed using repr(), while before you used print on the float directly, which uses str()

repr() and str() output of floating point numbers simply differs:

>>> lst = [0, 0.1, 0.2, 0.30000000000000004, 0.4, 0.5, 0.6, 0.7, 0.7999999999999999, 0.8999999999999999, 0.9999999999999999]
>>> print lst
[0, 0.1, 0.2, 0.30000000000000004, 0.4, 0.5, 0.6, 0.7, 0.7999999999999999, 0.8999999999999999, 0.9999999999999999]
>>> for elem in lst:
...     print elem
... 
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
>>> str(lst[3])
'0.3'
>>> repr(lst[3])
'0.30000000000000004'

repr() on a float produces a result that'll let you reproduce the same value accurately. str() rounds the floating point number for presentation.

answered Oct 21, 2014 at 17:00

Martijn Pieters

1.1m326 gold badges4.2k silver badges3.4k bronze badges

Sign up to request clarification or add additional context in comments.

4 Comments

WGS Over a year ago

Wondering, though... Does this mean repr() is subject to the same floating point representation inaccuracy as detailed here?

Martijn Pieters Over a year ago

@Nanashi: of course it is. All that repr() does is reveal the actual float value, not round it to a human readable version.

WGS Over a year ago

Wow, I didn't know that's how repr() works at all. This shed more light than an awful lot of other repr() vs str() tutorials I've read. Thanks!

Mark Ransom Over a year ago

@NullDev note that around Python 3.2, things changed so that for floating point numbers there's no difference between str and repr. I've edited the relevant link into the duplicates list.

Collectives™ on Stack Overflow

Last decimal digit precision changes in different call of same generator function [python] [duplicate]

1 Answer 1

4 Comments

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

4 Comments

Linked

Related