On Mon, May 11, 2009 at 3:06 AM, Eric Firing <ef...@ha...> wrote:
> C M wrote:
>>
>> On Mon, May 11, 2009 at 12:52 AM, Gökhan SEVER <gok...@gm...>
>> wrote:
>>>
>>> And the answer is:
>>>
>>> axis(xmin=..., xmax=...)
>>>
>>> Probably, that was a very easy question and no one wanted to answer :)
>>>
>>> Gökhan
>>>
>>>
>>> On Sun, May 10, 2009 at 4:12 PM, Gökhan SEVER <gok...@gm...>
>>> wrote:
>>>>
>>>> Hello,
>>>>
>>>> I overlay bunch of boxplots with mean values shown as stars on each
>>>> corresponding boxplot instance. (As could be seen in this image:
>>>> http://img216.imageshack.us/img216/7528/boxplot.png.
>>>>
>>>> There is a minor thing that affects the appearance of the figure. That
>>>> is
>>>> 1st and the last boxplots don't fit in the figure borders. How can I fix
>>>> this? Do you have any suggestions?
>>>>
>>>> Thank you.
>>>>
>>>> Gökhan
>>
>> What is the difference between doing that or using something like:
>>
>> axes.autoscale_view(tight=False, scalex=True, scaley=True)
>>
>> I am under the impression that if tight=False, the plot does not
>> autoscale to precisely the xlims, but leaves some margin. Is
>> that right?
>
> Not necessarily--there is no guaranteed margin, although making such a
> margin a settable parameter has been in mind for a long time. With
> tight=False, the autoscaling goes to the nearest ticks that include the data
> range, so depending on the data range and the tick intervals, there might be
> negligible margin.
>
> Eric
Useful to know, thanks.
Che
|
