Otacilio and I exchanged some more emails and have determined that this is due to a long-standing bug, that occurs rarely when drawing lines. In Ferret, all lines are drawn by calls to the graphics package to move to a location, or draw to a location. When the coordinates cross a plot boundary, the location of the intercept is found, so the line goes just up to that border.
When there are two segments in a row that are outside but near one of the corners, like this:
then we can get the kinds of stray lines Otacilio saw. In his case it was triggered by a particular combination of the plot commands and "go land". I have found a fix for this, which will be in the next release of Ferret.
Ansley
On 10/15/2011 7:59 AM, Otacilio Leandro wrote:
Hi Ansley, My Ferret version is: ferret -nojnl NOAA/PMEL TMAP FERRET v6.72 Linux 2.6.32-131.12.1.el6.x86_64 32-bit - 09/13/11 15-Oct-11 10:54 Trying to answer all your questions: As you suspected is the "go land" the cause of problem. That strange line apear after the plot is made. See test-1.gif without "go land" and test.gif with "go land". I saved just the fist timestep [l=1] of each file and seding attached. I'm already seding vento-test.jnl, an edited version of the 1st script, to read these new files. Thanks for the help. 2011/10/14 Ansley Manke <Ansley.B.Manke@xxxxxxxx>:Hi, Let's talk about this outside the list and then report back when we find an answer. Ferret's underlying graphics draw lines by issuing commands to "move to a location" or "draw to a location". Sometimes we see a bug where these are out of sync, so that the location information is corrupted. What version of Ferret are you running? I have seen these stray lines sometimes on plots that were related to "go land". As an experiment, does the plot change if you comment out the "go land"? If you put in a pause statement after each plot is drawn, before the next "set viewport", do you see the line being drawn upon finishing the plot? Or is it happening when the next plot is started? If you put in more pause statements, you might be able to see when it's occurring. Are your datasets large? If not, then perhaps you could send them to me, or a subset of them, and I could check into this further. Ansley On 10/14/2011 11:19 AM, Otacilio Leandro wrote:Hi Ferrets, I'm analising the climatology of wind at 200mb over South America. I'm having problems (se the attached figure) when using vectors and viewports. The figure was done corretly when i used the /flow rather than /vector. Any sugestions?? Tks