Re: Question about PlotRange
- To: mathgroup at smc.vnet.net
- Subject: [mg129506] Re: Question about PlotRange
- From: James Stein <mathgroup at stein.org>
- Date: Sat, 19 Jan 2013 01:13:29 -0500 (EST)
- Delivered-to: l-mathgroup@mail-archive0.wolfram.com
- Delivered-to: l-mathgroup@wolfram.com
- Delivered-to: mathgroup-newout@smc.vnet.net
- Delivered-to: mathgroup-newsend@smc.vnet.net
- References: <20130118054932.C5DB167D9@smc.vnet.net>
Bob Hanlon: Many thanks for your excellent workarounds.
Bill Rowe: you say the second chart is not what you expected. Not what
I expected either; that was the thrust of my query.
I expected that when a value was beyond the plot range, a portion of
the fill line would nevertheless appear -- exactly that part of the
fill line that IS in the plot range. And this is, in fact, how
Mathematica displayed the (two) extreme values (one positive, one
negative).
If all points out side the range were neither plotted nor filled, I
could understand that (I see what the underlying algorithm might be).
And if all points outside the range were not plotted, but were filled
(as much as could be shown), I could understand that too(and imagine
the underlying algorithm). But I was surprised -- you too, evidently
-- that Mathematica does BOTH, and I can't imagine what algorithm or
interpretation it is implementing. Pending an explanation, I deem it a
bug.
One possible interpretation of PlotRange is this: Imagine the graph
drawn in full; and then imagine the region of the graph what is within
the PlotRange; this second imagined region is what I expected ListPlot
to display.
PS. I see where I misled you. My original post said "My desire is for
all 27 values to appear on the second chart." What I should have said
was "My desire is to see some visible manifestation of all 27 values"
-- that was, in fact, the purpose of using "Filling->0". Even for
out-of-range points, their filling would not (completely) be out of
range.
On Thu, Jan 17, 2013 at 9:49 PM, Bill Rowe <readnews at sbcglobal.net> wrote:
> On 1/16/13 at 1:42 AM, james at stein.org (James Stein) wrote:
>
>>I am mystified by the two charts produced by this expression:
>
>>Table [
>>ListPlot [ Table [ x, { x, -.13, .13, .01 } ] , Axes -> True,
>>Filling -> 0, FillingStyle -> { Red, Darker [ Green ] } , PlotRange
>>-> { Automatic, yRange }
>>], { yRange, { Full, { -0.1, 0.06} } } ]
>
>>If 'PlotRange' allows 'Full' for the y-axis, all data is plotted as
>>expected, but if 'PlotRange' restricts extreme values of y, *some*
>>extreme values have vanished, but *not* the *most* extreme values.
>
>>My desire is for all 27 values to appear on the second chart. (In
>>more realistic situations, the visual 'Filling' is important.) How
>>is this to be achieved?
>
> I probably don't understand what you want with the second chart.
> And I do admit the graphic I see in version 9 running on OS X
> 10.8.2 isn't quite what I would have expected. But I don't see
> why you are expecting the second plot to show all 27 values.
>
> The last 8 values are:
>
> In[14]:= Table[x, {x, -.13, .13, .01}][[-7 ;;]]
>
> Out[14]= {0.07,0.08,0.09,0.1,0.11,0.12,0.13}
>
> all of which are greater than the maximum value (0.06) you
> provide for PlotRange in the second plot. Clearly the cannot be
> plotted with that plot range specification.
>
> As for the most extreme positive value (0.13) the value is not
> plotted. All that is shown is a filling line truncated at 0.06,
> the limit specified by the plot range you provided. What I find
> a bit surprising and don't understand is why only the first and
> last values are shown with a truncated filling line while other
> values outside the plot range show no similar truncated filling
> line. I would have expected either all values outside to the
> plot range to either be shown with a truncated filling line or
> not shown at all.
>
>
- References:
- Re: Question about PlotRange
- From: Bill Rowe <readnews@sbcglobal.net>
- Re: Question about PlotRange