matplotlib-users Mailing List for matplotlib

On Thursday 25 October 2007 03:17:05 pm David D Clark wrote:
> Hello folks,
>
> I have a Gaussian peak centered on zero.  When the plot is drawn on
> screen with the GTKAgg backend, the peak is shown in the correct place.
>  When I either click on the disk icon and save the plot as eps or use
> savefig, the peak is shifted to the left of zero by several x axis units.
>
> I am using Python 2.5 with matplotlib 0.90.1 and numpy 1.0.3. and scipy
> 0.5.1 in my script.  These are running on my Fedora 7 box. I compiled
> and installed mpl 0.90.1 this morning because I noticed this behavior
> with python-matplotlib-0.90.0 that I had installed with yum.  I also see
> this behavior if I run my script from the command line and from ipython
> 0.8.1.
>
> When I save using PNG, the peak is in the correct place.  My Mac also
> saves both EPS and PNG correctly.
>
> Any thoughts?

Maybe you could post a small example that demonstrates the problem. I dont see 
any reason why eps output should be different depending on your OS, unless it 
is a problem with your viewer, or you are distilling your postscript output 
and it is a problem with ghostscript.

Darren

Jeff -

My data set is actually dynamically generated by a program called  
ShakeMap.  It's a 2D grid, with an extent usually about 600  
kilometers on a side, centered wherever the earthquake happened to  
be.  The ShakeMap program does not know or care that some of the data  
may be under water, but for display purposed, I do!  The grid is also  
in a geographic projection (latitude/longitude coordinates assumed to  
be cartesian).

So in this test instance (on a data set near Taiwan), my map width is  
about 5.91 degrees longitude, and my height is about 5.5 degrees  
latitude.

If I set xoffset=-0.01*5.91, I get -0.05.  This is not noticeably  
different than the default.

Is the problem that my dataset is not projected?

--Mike
On Nov 2, 2007, at 1:33 PM, Jeff Whitaker wrote:

> Michael Hearne wrote:
>> Jeff - I looked at that example file, and I think there's a big  
>> difference - your etopo base data set is global, and you can plot  
>> over the data in the oceans by setting the mask on all pixels less  
>> than zero.
>>
>> My dataset (a map of earthquake shaking) is not global, and  
>> actually has NO missing data.   I think I need a way to "clip" the  
>> data by the land mask - that is, find all of the pixels that are  
>> NOT on land, and then mask them off.
> Mike:
>
> If it's not global, is it just defined for land points?  If so, it  
> can't be a 2-D grid, so you won't be able to plot it with imshow  
> anyway.  Can you explain the structure of the data?
>>
>> Is there an easy way to do this with matplotlib/basemap tools?
> Not really.  You'll have to define a sea mask for your grid and use  
> that the create a masked array.  There is a land-sea mask dataset  
> included in basemap, but it may not match the resolution of your grid.
>
>>
>> Regarding my other issue - I used my script to test x/y offset  
>> values: [0.05,0.1,0.5,1.0,10] and couldn't see any difference.   
>> I'd be more than happy to provide test output, or debugging  
>> information...
>>
>> Just to be clear - these offsets are supposed to move the meridian  
>> and/or parallel labels around with respect to the map edge?  My  
>> actual goal is to get the labels inside the edge of the map (I  
>> tried negative numbers to accomplish this, to no effect.)
> You  need to define an offset as a fraction of the map width - the  
> numbers you are giving are too small to notice any difference.  As  
> I said before, try something like -0.01*(m.max-m.min).
>>
>> On a positive note, I _can_ make solid lines!
>
> Good!
>
> -Jeff
>
> -- 
> Jeffrey S. Whitaker         Phone  : (303)497-6313
> Meteorologist               FAX    : (303)497-6449
> NOAA/OAR/PSD  R/PSD1        Email  : Jef...@no...
> 325 Broadway                Office : Skaggs Research Cntr 1D-124
> Boulder, CO, USA 80303-3328 Web    : http://tinyurl.com/5telg




------------------------------------------------------
Michael Hearne
mh...@us...
(303) 273-8620
USGS National Earthquake Information Center
1711 Illinois St. Golden CO 80401
Senior Software Engineer
Synergetics, Inc.
------------------------------------------------------

On Friday 02 November 2007 15:10:15 Michael Hearne wrote:

> Is there an easy way to do this with matplotlib/basemap tools?

I'm afraid that not.
Your problem is that you need to find for each cell of your gridded data 
whether you are on land or in an ocean, and mask the corresponding cell in 
the latter case. Such a manipulation of geometry is not available in 
matplotlib (yet, and will it ever be?). 
You could of course start writing your tests. However, that's exactly what OGR 
is designed to do, and the couple of functions I sent earlier can be quite 
helpful. Once again, OGR is part of the GDAL suite. 

Michael Hearne wrote:
> Jeff - I looked at that example file, and I think there's a big 
> difference - your etopo base data set is global, and you can plot over 
> the data in the oceans by setting the mask on all pixels less than zero.
>
> My dataset (a map of earthquake shaking) is not global, and actually 
> has NO missing data.   I think I need a way to "clip" the data by the 
> land mask - that is, find all of the pixels that are NOT on land, and 
> then mask them off. 
Mike:

If it's not global, is it just defined for land points?  If so, it can't 
be a 2-D grid, so you won't be able to plot it with imshow anyway.  Can 
you explain the structure of the data?
>
> Is there an easy way to do this with matplotlib/basemap tools?
Not really.  You'll have to define a sea mask for your grid and use that 
the create a masked array.  There is a land-sea mask dataset included in 
basemap, but it may not match the resolution of your grid.

>
> Regarding my other issue - I used my script to test x/y offset values: 
> [0.05,0.1,0.5,1.0,10] and couldn't see any difference.  I'd be more 
> than happy to provide test output, or debugging information...
>
> Just to be clear - these offsets are supposed to move the meridian 
> and/or parallel labels around with respect to the map edge?  My actual 
> goal is to get the labels inside the edge of the map (I tried negative 
> numbers to accomplish this, to no effect.)
You  need to define an offset as a fraction of the map width - the 
numbers you are giving are too small to notice any difference.  As I 
said before, try something like -0.01*(m.max-m.min).
>
> On a positive note, I _can_ make solid lines!

Good!

-Jeff

-- 
Jeffrey S. Whitaker         Phone  : (303)497-6313
Meteorologist               FAX    : (303)497-6449
NOAA/OAR/PSD  R/PSD1        Email  : Jef...@no...
325 Broadway                Office : Skaggs Research Cntr 1D-124
Boulder, CO, USA 80303-3328 Web    : http://tinyurl.com/5telg

Jeff - I looked at that example file, and I think there's a big  
difference - your etopo base data set is global, and you can plot  
over the data in the oceans by setting the mask on all pixels less  
than zero.

My dataset (a map of earthquake shaking) is not global, and actually  
has NO missing data.   I think I need a way to "clip" the data by the  
land mask - that is, find all of the pixels that are NOT on land, and  
then mask them off.

Is there an easy way to do this with matplotlib/basemap tools?

Regarding my other issue - I used my script to test x/y offset  
values: [0.05,0.1,0.5,1.0,10] and couldn't see any difference.  I'd  
be more than happy to provide test output, or debugging information...

Just to be clear - these offsets are supposed to move the meridian  
and/or parallel labels around with respect to the map edge?  My  
actual goal is to get the labels inside the edge of the map (I tried  
negative numbers to accomplish this, to no effect.)

On a positive note, I _can_ make solid lines!

Thanks for all of your help,

Mike


On Nov 2, 2007, at 11:47 AM, Jeff Whitaker wrote:

> Michael Hearne wrote:
>> I have two questions:
>>
>> 1)  The fillcontinents() method has a zorder keyword parameter.   
>> Is this supposed to work with imshow()?  I have the latest tarball  
>> from the website, and I can't get my image to paint on top of the  
>> continents.
>>
>> 2) Has anyone figured out a way to make an _ocean_ mask?  I need  
>> my map to look like this
>> ( http://earthquake.usgs.gov/eqcenter/pager/us/2007itah/us/5/ 
>> shakemap.600.jpg)
>>
>> , where the image data (the yellow and orange bits) that extends  
>> out beyond the land boundary is masked by a blue ocean.
>>
>> Thanks,
>>
>> Mike
> Mike:  You can set the background color for the axes instance to  
> blue, and then draw the filled continents over it.
>
> You'll also have to make your image a masked array, and set the  
> missing values to be transparent.
>
> The basemap examples plotmap_masked.py shows how to do both of  
> these things.
>
> -Jeff
> -- 
> Jeffrey S. Whitaker         Phone  : (303)497-6313
> Meteorologist               FAX    : (303)497-6449
> NOAA/OAR/PSD  R/PSD1        Email  : Jef...@no...
> 325 Broadway                Office : Skaggs Research Cntr 1D-124
> Boulder, CO, USA 80303-3328 Web    : http://tinyurl.com/5telg




------------------------------------------------------
Michael Hearne
mh...@us...
(303) 273-8620
USGS National Earthquake Information Center
1711 Illinois St. Golden CO 80401
Senior Software Engineer
Synergetics, Inc.
------------------------------------------------------

Michael Hearne wrote:
> Thanks for the reponses so far, but I lied.  I have more questions :)  
>
> 1)  How do I get drawmeridians() and drawparallels() to draw solid 
> (i.e., not dashed) lines?  setting linestyle='-' does not seem to work.
Mike:  From the docstring:

 "dashes - dash pattern for parallels (default [1,1], i.e. 1 pixel on, 1 
pixel off)."

So use dashes=[1,0] to get solid lines.

> 2) What are appropriate values for the xoffset and yoffset keyword 
> parameters for the above functions?  I've tried several values, even 
> altering them by a few orders of magnitude, and I can't see any 
> difference. 
"xoffset - label offset from edge of map in x-direction
  (default is 0.01 times width of map in map projection coordinates).
 yoffset - label offset from edge of map in y-direction
  (default is 0.01 times height of map in map projection coordinates)."

So, the default is 0.01*(m.xmax-m.xmin).   Making it 0.1*(m.xmax-m.xmin) 
makes the labels much farther out for me.

-Jeff


-- 
Jeffrey S. Whitaker         Phone  : (303)497-6313
Meteorologist               FAX    : (303)497-6449
NOAA/OAR/PSD  R/PSD1        Email  : Jef...@no...
325 Broadway                Office : Skaggs Research Cntr 1D-124
Boulder, CO, USA 80303-3328 Web    : http://tinyurl.com/5telg

Michael Hearne wrote:
> I have two questions:
>
> 1)  The fillcontinents() method has a zorder keyword parameter.  Is 
> this supposed to work with imshow()?  I have the latest tarball from 
> the website, and I can't get my image to paint on top of the continents.
>
> 2) Has anyone figured out a way to make an _ocean_ mask?  I need my 
> map to look like this 
>
> ( 
> http://earthquake.usgs.gov/eqcenter/pager/us/2007itah/us/5/shakemap.600.jpg)  
>
>
> , where the image data (the yellow and orange bits) that extends out 
> beyond the land boundary is masked by a blue ocean.
>
> Thanks,
>
> Mike
Mike:  You can set the background color for the axes instance to blue, and then draw the filled continents over it.

You'll also have to make your image a masked array, and set the missing values to be transparent.

The basemap examples plotmap_masked.py shows how to do both of these things.

-Jeff
-- 
Jeffrey S. Whitaker         Phone  : (303)497-6313
Meteorologist               FAX    : (303)497-6449
NOAA/OAR/PSD  R/PSD1        Email  : Jef...@no...
325 Broadway                Office : Skaggs Research Cntr 1D-124
Boulder, CO, USA 80303-3328 Web    : http://tinyurl.com/5telg

Pierre - That's great.  As soon as I get the zorder problem figured  
out (can't have land AND ocean masks on!) , I'll give your solution a  
try.

One question regarding - I found this page:
http://trac.osgeo.org/gdal/wiki/GdalOgrInPython

but it wasn't immediately apparent how to _get_ it.  Do I have to  
download/build ogr first, then install the python bindings?

--Mike
On Nov 2, 2007, at 10:06 AM, Pierre GM wrote:

> On Friday 02 November 2007 11:51:55 Michael Hearne wrote:
>> 2) Has anyone figured out a way to make an _ocean_ mask?  I need my
>> map to look like this
>
> Assuming that you have gdal installed, you can use this set of  
> functions.
> Basically, you use OGR to compute the differences between the  
> background and
> the land polygons... Works fine for my applications (I interpolate  
> some
> rainfall fields over AL/GA/FL, and need to hide the interpolated  
> values in
> the Gulf of Mexico), but might still be a tad buggy. Please don't  
> hesittate
> to contact me if you need help with that.
>
>
>
> #####----------------------------------------------------------------- 
> ---------
> #---- --- OGR conversion ---
> #####----------------------------------------------------------------- 
> ---------
> import ogr
>
> def polygon_to_geometry(polygon):
>     """Creates a new ogr.Geometry object from a matplolib.Polygon."""
>     if not isinstance(polygon, Polygon):
>         raise TypeError, "The input data should be a valid Polygon  
> object!"
>     listvert = ["%s %s" % (x,y)  for (x,y) in polygon.get_verts()]
>     listvert.append(listvert[0])
>     return ogr.CreateGeometryFromWkt("POLYGON ((%s))" % ','.join 
> (listvert))
>
> #--------------------------------------------------------------------- 
> --------
> #---- OGR to matplotlib ---
> #--------------------------------------------------------------------- 
> --------
>
> def _geometry_to_vertices(geometry):
>     """Creates a list of vertices (x,y) from the current geometry."""
>     verts = []
>     for nl in range(geometry.GetGeometryCount()):
>         line = geometry.GetGeometryRef(nl)
>         points = [(line.GetX(i), line.GetY(i)) for i in
> range(line.GetPointCount())]
>         verts.append(points)
>     return verts
>
>
> def geometry_to_vertices(geometry):
>     """Creates lists of vertices (x,y) from the current geometry."""
>     if not isinstance(geometry, ogr.Geometry):
>         raise TypeError, "The input data should be a valid  
> ogr.Geometry
> object!"
>     vertices = []
>     #
>     if geometry.GetGeometryType() == ogr.wkbPolygon:
>         vertices.extend(_geometry_to_vertices(geometry))
>     elif geometry.GetGeometryType() == ogr.wkbMultiPolygon:
>         for np in range(geometry.GetGeometryCount()):
>             poly = geometry.GetGeometryRef(np)
>             vertices.extend(_geometry_to_vertices(poly))
>     return vertices
>
> #####----------------------------------------------------------------- 
> ---------
> #---- --- Add-ons
> #####----------------------------------------------------------------- 
> ---------
>
> def fillwaterbodies(basemap, color='blue', inlands=True, ax=None,
> zorder=None):
>     """Fills the water bodies with color.
>     If inlands is True, inland water bodies are also filled.
>
> :Inputs:
>     basemap : Basemap
>         The basemap on which to fill.
>     color : string/RGBA tuple *['blue']*
>         Filling color
>     inlands : boolean *[True]*
>         Whether inland water bodies should be filled.
>     ax : Axes instance *[None]*
>         Axe on which to plot. If None, the current axe is selected.
>     zorder : integer *[None]*
>         zorder of the water bodies polygons.
>     """
>     if not isinstance(basemap, Basemap):
>         raise TypeError, "The input basemap should be a valid  
> Basemap object!"
>     #
>     if ax is None and basemap.ax is None:
>         try:
>             ax = pylab.gca()
>         except:
>             import pylab
>             ax = pylab.gca()
>     #
>     coastpolygons = basemap.coastpolygons
>     coastpolygontypes = basemap.coastpolygontypes
>     (llx, lly) = (basemap.llcrnrx, basemap.llcrnry)
>     (urx, ury) = (basemap.urcrnrx,basemap.urcrnry)
>     background = Polygon([(llx, lly), (urx, lly), (urx, ury), (llx,  
> ury)])
>     #
>     ogr_background = polygon_to_geometry(background)
>     inland_polygons = []
>     #
>     for (poly, polytype) in zip(coastpolygons, coastpolygontypes):
>         if polytype != 2:
>             verts = ["%s %s" % (x,y) for (x,y) in zip(*poly)]
>             ogr_poly = ogr.CreateGeometryFromWkt("POLYGON
> ((%s))" % ','.join(verts))
>             ogr_background = ogr_background.Difference(ogr_poly)
>         else:
>             inland_polygons.append(Polygon(zip(*poly),
>                                            facecolor=color,
>                                            edgecolor=color,
>                                            linewidth=0))
>     #
>     background = geometry_to_vertices(ogr_background)
>     for xy in background:
>         patch = Polygon(xy, facecolor=color, edgecolor=color,  
> linewidth=0)
>         if zorder is not None:
>             patch.set_zorder(zorder)
>         ax.add_patch(patch)
>     #
>     if inlands:
>         for poly in inland_polygons:
>             if zorder is not None:
>                 poly.set_zorder(zorder)
>             ax.add_patch(poly)
>     basemap.set_axes_limits(ax=ax)
>     return
>
> ---------------------------------------------------------------------- 
> ---
> This SF.net email is sponsored by: Splunk Inc.
> Still grepping through log files to find problems?  Stop.
> Now Search log events and configuration files using AJAX and a  
> browser.
> Download your FREE copy of Splunk now >> http://get.splunk.com/
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users




------------------------------------------------------
Michael Hearne
mh...@us...
(303) 273-8620
USGS National Earthquake Information Center
1711 Illinois St. Golden CO 80401
Senior Software Engineer
Synergetics, Inc.
------------------------------------------------------

On Friday 02 November 2007 11:51:55 Michael Hearne wrote:
> 2) Has anyone figured out a way to make an _ocean_ mask?  I need my
> map to look like this

Assuming that you have gdal installed, you can use this set of functions. 
Basically, you use OGR to compute the differences between the background and 
the land polygons... Works fine for my applications (I interpolate some 
rainfall fields over AL/GA/FL, and need to hide the interpolated values in 
the Gulf of Mexico), but might still be a tad buggy. Please don't hesittate 
to contact me if you need help with that.



#####--------------------------------------------------------------------------
#---- --- OGR conversion ---
#####--------------------------------------------------------------------------
import ogr

def polygon_to_geometry(polygon):
    """Creates a new ogr.Geometry object from a matplolib.Polygon."""
    if not isinstance(polygon, Polygon):
        raise TypeError, "The input data should be a valid Polygon object!"
    listvert = ["%s %s" % (x,y)  for (x,y) in polygon.get_verts()]
    listvert.append(listvert[0])
    return ogr.CreateGeometryFromWkt("POLYGON ((%s))" % ','.join(listvert))

#-----------------------------------------------------------------------------
#---- OGR to matplotlib ---
#-----------------------------------------------------------------------------

def _geometry_to_vertices(geometry):
    """Creates a list of vertices (x,y) from the current geometry."""
    verts = []
    for nl in range(geometry.GetGeometryCount()):
        line = geometry.GetGeometryRef(nl)
        points = [(line.GetX(i), line.GetY(i)) for i in 
range(line.GetPointCount())]
        verts.append(points)
    return verts
    

def geometry_to_vertices(geometry):
    """Creates lists of vertices (x,y) from the current geometry."""
    if not isinstance(geometry, ogr.Geometry):
        raise TypeError, "The input data should be a valid ogr.Geometry 
object!"
    vertices = []
    #
    if geometry.GetGeometryType() == ogr.wkbPolygon:
        vertices.extend(_geometry_to_vertices(geometry))
    elif geometry.GetGeometryType() == ogr.wkbMultiPolygon:
        for np in range(geometry.GetGeometryCount()):
            poly = geometry.GetGeometryRef(np)
            vertices.extend(_geometry_to_vertices(poly))
    return vertices

#####--------------------------------------------------------------------------
#---- --- Add-ons
#####--------------------------------------------------------------------------

def fillwaterbodies(basemap, color='blue', inlands=True, ax=None, 
zorder=None): 
    """Fills the water bodies with color.
    If inlands is True, inland water bodies are also filled.
    
:Inputs:
    basemap : Basemap
        The basemap on which to fill.
    color : string/RGBA tuple *['blue']*
        Filling color
    inlands : boolean *[True]*
        Whether inland water bodies should be filled.
    ax : Axes instance *[None]*
        Axe on which to plot. If None, the current axe is selected.
    zorder : integer *[None]*
        zorder of the water bodies polygons.    
    """   
    if not isinstance(basemap, Basemap):
        raise TypeError, "The input basemap should be a valid Basemap object!"
    #
    if ax is None and basemap.ax is None:
        try:
            ax = pylab.gca()
        except:
            import pylab
            ax = pylab.gca()
    #
    coastpolygons = basemap.coastpolygons
    coastpolygontypes = basemap.coastpolygontypes
    (llx, lly) = (basemap.llcrnrx, basemap.llcrnry)
    (urx, ury) = (basemap.urcrnrx,basemap.urcrnry)
    background = Polygon([(llx, lly), (urx, lly), (urx, ury), (llx, ury)])
    #
    ogr_background = polygon_to_geometry(background)
    inland_polygons = []
    #
    for (poly, polytype) in zip(coastpolygons, coastpolygontypes):
        if polytype != 2:
            verts = ["%s %s" % (x,y) for (x,y) in zip(*poly)]
            ogr_poly = ogr.CreateGeometryFromWkt("POLYGON 
((%s))" % ','.join(verts))
            ogr_background = ogr_background.Difference(ogr_poly)
        else:
            inland_polygons.append(Polygon(zip(*poly), 
                                           facecolor=color,
                                           edgecolor=color, 
                                           linewidth=0))
    #
    background = geometry_to_vertices(ogr_background)
    for xy in background:
        patch = Polygon(xy, facecolor=color, edgecolor=color, linewidth=0)
        if zorder is not None:
            patch.set_zorder(zorder)
        ax.add_patch(patch)
    #        
    if inlands:
        for poly in inland_polygons:
            if zorder is not None:
                poly.set_zorder(zorder)
            ax.add_patch(poly)
    basemap.set_axes_limits(ax=ax)
    return

I have two questions:

1)  The fillcontinents() method has a zorder keyword parameter.  Is  
this supposed to work with imshow()?  I have the latest tarball from  
the website, and I can't get my image to paint on top of the continents.

2) Has anyone figured out a way to make an _ocean_ mask?  I need my  
map to look like this

(http://earthquake.usgs.gov/eqcenter/pager/us/2007itah/us/5/shakemap. 
600.jpg)

, where the image data (the yellow and orange bits) that extends out  
beyond the land boundary is masked by a blue ocean.

Thanks,

Mike

On Thu, 1 Nov 2007
Alan G Isaac <ai...@am...> wrote:
> <URL:http://www.mail-archive.com/mat...@li.../msg01221.html>
>
>hth,

Ah, I sould be able to modify that to do what I need but it's to bad
there isn't a more elegant way than remaping the data and faking the
labels.

Thanks for the quick help.

-- 
Jed Frechette
jdfrechette.alturl.com