matplotlib-users Mailing List for matplotlib (Page 12)

--- On Tue, 11/2/10, Benjamin Root <ben...@ou...> wrote:
I
 have personally seen significant progress in this area, but there are a
 few backends that aren't quite right (MacOSX backend, I believe?).

As far as I know, show() is working correctly in the MacOSX backend. If you encountered a case where the backend doesn't do the right thing, can you show a concrete example?

Best,
--Michiel.




      

On 11/02/2010 03:18 PM, Jae-Joon Lee wrote:
> A quick (and not safe) way w/ mpl v1.0 is,
>
>      ax = plt.subplot(111)
>      ax.plot(np.arange(3))
>
>      ax.set_xticks([0, 0.5, 1., 1.5, 2.])
>
>      mytick = ax.xaxis.majorTicks[2]
>      mytick._apply_params(tickdir="out")
>
> I don't think there is a way to do this only using public apis.
> I myself actually prefer to create a separate axis (or axes), but this
> will be more trickier than above solution.

Each tick is a Line2D consisting of a single marker, so it is possible 
to change the marker after the tick has been created:

ax = plt.subplot(1,1,1)
ticks = ax.xaxis.get_majorticklines()
mytick = ticks[3]
mytick.set_marker(2)
plt.draw()

This will flip the first tick on the top border (the one for 0.2).
left, right, up, down are 0, 1, 2, 3.

Eric



>
> Regards,
>
> -JJ
>
>
>
>
> On Tue, Nov 2, 2010 at 4:22 AM, Bartosz Telenczuk
> <b.t...@bi...>  wrote:
>> Hi all,
>>
>> Is it possible to set direction (in or out) individually for each tick. I know about the rc setting ("(x/y)tick.direction") , but I need a finer control over the ticks.
>>
>> Thanks,
>>
>> Bartek
>>
>> ------------------------------------------------------------------------------
>> Nokia and AT&T present the 2010 Calling All Innovators-North America contest
>> Create new apps&  games for the Nokia N8 for consumers in  U.S. and Canada
>> $10 million total in prizes - $4M cash, 500 devices, nearly $6M in marketing
>> Develop with Nokia Qt SDK, Web Runtime, or Java and Publish to Ovi Store
>> http://p.sf.net/sfu/nokia-dev2dev
>> _______________________________________________
>> Matplotlib-users mailing list
>> Mat...@li...
>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>>
>
> ------------------------------------------------------------------------------
> Achieve Improved Network Security with IP and DNS Reputation.
> Defend against bad network traffic, including botnets, malware,
> phishing sites, and compromised hosts - saving your company time,
> money, and embarrassment.   Learn More!
> http://p.sf.net/sfu/hpdev2dev-nov
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

I am trying to use matplotlib (for the first time) to graph the address space 
usage of an application against time.  The data is written to a log file by 
trace statements throughout the source code of the application.  The trace 
statements contain the current address space usage as well as a timer value 
with millisecond granularity.

My data in the y-axis (address space usage) is fairly uniform (0-2000 MB 
values), but my data in the x-axis (the time at which the the trace statements 
were executed) is highly clustered.  For example, I have approximately 150 
data points over a 5 minute run, but some of the data points are only 10ms 
apart.

I would like to annotate each point on the graph with the line number in the 
log file so that the user can look up what was happening at that point.  I have 
succeeded, but the graph isn't readable because there is so much overlap in 
the points.

Is there a standard way that people display data like this?  I don't really 
like the idea of equally spacing all of the points along the x-axis because 
you lose the understanding of the timing.  One idea I had was to have some 
sort of vertical break in the graph at areas where there was a long gap 
without a data point, but I have no idea whether it's possible to implement 
something like that in matplotlib.

The output format hasn't been strictly specified, so if you have any ideas of 
how I can produce a useful graph, I would be happy to hear them.

Thanks,
Dave

Hi,

Fantastic, path_effect is exactly what I'm looking for. Thanks!

-Niko

On 11/02/2010 10:16 PM, Jae-Joon Lee wrote:
> You may use annotate with which you can specify offsets.
> 
> http://matplotlib.sourceforge.net/api/pyplot_api.html?highlight=annotate#matplotlib.pyplot.annotate
> 
> Or, you may consider to use the path_effect (available w/ v1.0).
> 
> http://matplotlib.sourceforge.net/examples/pylab_examples/patheffect_demo.html
> 
> Regards,
> 
> -JJ
> 
> 
> On Wed, Nov 3, 2010 at 10:44 AM, Nikolaus Rath <Nik...@ra...> wrote:
>> Hello,
>>
>> I am placing a text object in a place where the background is sometimes
>> dark and sometimes light, so the text is hard to see. Adding a
>> background to the text object itself looks a bit ugly, so I am wondering
>> if there is a way to add a shadow to the text itself.
>>
>> What's the best way to do this?
>>
>> I thought about redrawing the same text just offset by a fraction of the
>> text size, but I can't quite figure out how to do the placement
>> properly. My rough idea is the get the position of the text in pixel
>> coordinates, add an offset to it and then position it with display
>> coordinates as well. But how do I get the display coordinates of
>> something that I placed in axes coordinates? And how to I determine how
>> many pixels I need to offset without knowing the dpi?
>>
>>
>> Best,
>>
>>   -Nikolaus
>>
>> --
>>  »Time flies like an arrow, fruit flies like a Banana.«
>>
>>  PGP fingerprint: 5B93 61F8 4EA2 E279 ABF6  02CF A9AD B7F8 AE4E 425C
>>
>> ------------------------------------------------------------------------------
>> Achieve Improved Network Security with IP and DNS Reputation.
>> Defend against bad network traffic, including botnets, malware,
>> phishing sites, and compromised hosts - saving your company time,
>> money, and embarrassment.   Learn More!
>> http://p.sf.net/sfu/hpdev2dev-nov
>> _______________________________________________
>> Matplotlib-users mailing list
>> Mat...@li...
>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>>


   -Nikolaus

-- 
 »Time flies like an arrow, fruit flies like a Banana.«

  PGP fingerprint: 5B93 61F8 4EA2 E279 ABF6  02CF A9AD B7F8 AE4E 425C

For whatever it's worth, after a lot of wrangling, I think I solved most of
my problems (though perhaps not in the most efficient way).

In case anyone else is looking for similar functionality, here's a callback
function that will autowrap text objects to the inside of the axis they're
plotted in, and should handle any font, rotation, etc that you throw at it.
(The previous version had a lot of bugs).

Hope someone finds it useful, at any rate...
-Joe
[image: E2G1j.png]

import matplotlib.pyplot as plt

def main():
    """Draw some very long strings on a figure and have the auto-wrapped
    to the axis boundaries. Try resizing the figure!!"""

    fig = plt.figure()
    plt.axis([0, 10, 0, 10])

    t = "This is a really long string that I'd rather have wrapped so"\
        " that it doesn't go outside of the figure, but if it's long"\
        " enough it will go off the top or bottom!"

    t2 = r"Furthermore, if I put mathtext in here, it won't mutilate it,"\
        " but will treat it like a really long word. For example: "\
        r"$\frac{\sigma}{\gamma} - e^{\theta \pm 5}$ won't be mangled!"

    plt.text(5, 10, t2, size=14, ha='center', va='top', family='monospace')
    plt.text(3, 0, t, family='serif', style='italic', ha='right')

    plt.text(4, 1, t, ha='left', family='Times New Roman', rotation=15)
    plt.text(5, 3.5, t, ha='right', rotation=-15)

    plt.title("This is a really long title that I want to have wrapped so"\
             r" it does not go outside the axis boundaries", ha='center')

    # All we do to autowrap everything  is connect a callback function...
    fig.canvas.mpl_connect('draw_event', on_draw)

    plt.show()

def on_draw(event):
    """Auto-wraps all text objects in a figure at draw-time"""
    import matplotlib as mpl
    fig = event.canvas.figure

    # Cycle through all artists in all the axes in the figure
    for ax in fig.axes:
        for artist in ax.get_children():
            # If it's a text artist, wrap it...
            if isinstance(artist, mpl.text.Text):
                autowrap_text(artist, event.renderer)

    # Temporarily disconnect any callbacks to the draw event...
    # (To avoid recursion)
    func_handles = fig.canvas.callbacks.callbacks[event.name]
    fig.canvas.callbacks.callbacks[event.name] = {}
    # Re-draw the figure..
    fig.canvas.draw()
    # Reset the draw event callbacks
    fig.canvas.callbacks.callbacks[event.name] = func_handles

def autowrap_text(textobj, renderer):
    """Wraps the given matplotlib text object so that it doesn't exceed the
    boundaries of the axis it is plotted in."""
    # Get the starting position of the text in pixels...
    x0, y0 = textobj.get_transform().transform(textobj.get_position())
    # Get the extents of the current axis in pixels...
    clip = textobj.get_axes().get_window_extent()
    # Set the text to rotate about the left edge (nonsensical otherwise)
    textobj.set_rotation_mode('anchor')

    # Get the amount of space in the direction of rotation to the left and
    # right of x0, y0 (left and right are relative to the rotation)
    rotation = textobj.get_rotation()
    right_space = min_dist_inside((x0, y0), rotation, clip)
    left_space = min_dist_inside((x0, y0), rotation - 180, clip)

    # Use either the left or right distance depending on the h-alignment.
    alignment = textobj.get_horizontalalignment()
    if alignment is 'left':
        new_width = right_space
    elif alignment is 'right':
        new_width = left_space
    else:
        new_width = 2 * min(left_space, right_space)

    # Convert to characters with a minimum width of 1 character
    wrap_width = max(1, new_width // pixels_per_char(textobj))
    try:
        wrapped_text = safewrap(textobj.get_text(), wrap_width)
    except TypeError:
        # This appears to be a single word
        wrapped_text = textobj.get_text()
    textobj.set_text(wrapped_text)

def min_dist_inside(point, rotation, box):
    """Gets the space in a given direction from "point" to the boundaries
    of "box" (where box is an object with x0, y0, x1, & y1 attributes,
    point is a tuple of x,y, and rotation is the angle in degrees)"""
    from math import sin, cos, radians
    x0, y0 = point
    rotation = radians(rotation)
    distances = []
    threshold = 0.0001
    if cos(rotation) > threshold:
        # Intersects the right axis
        distances.append((box.x1 - x0) / cos(rotation))
    if cos(rotation) < -threshold:
        # Intersects the left axis
        distances.append((box.x0 - x0) / cos(rotation))
    if sin(rotation) > threshold:
        # Intersects the top axis
        distances.append((box.y1 - y0) / sin(rotation))
    if sin(rotation) < -threshold:
        # Intersects the bottom axis
        distances.append((box.y0 - y0) / sin(rotation))
    return min(distances)

def pixels_per_char(textobj):
    """Determines the average width of a character of the given textobj
    by drawing a test string and calculating it's length"""
    test_text = 'Try something like a test'
    orig_text = textobj.get_text()
    textobj.set_text(test_text)
    width = textobj.get_window_extent().width
    textobj.set_text(orig_text)
    return width / len(test_text)

def safewrap(text, width):
    """Wraps text, but avoids putting linebreaks in tex strings"""
    import textwrap
    # If it's not a tex string, just wrap it as usual...
    if '$' not in text:
        return textwrap.fill(text, width)

    # Tex segments will be inside two "$"'s, so we want the odd items
    segments = text.split('$')
    tex = segments[1::2]

    # Temporarily replace spaces and dashes inside tex segments so that
    # they will be treated as long words by textwrap...
    segments[1::2] = [x.replace(' ','').replace('-','') for x in tex]
    # Rejoin the temp tex strings with the rest of the text and wrap it
    temp_text = '$'.join(segments)
    wrapped = textwrap.fill(temp_text, width, break_long_words=False)

    # Put the original tex strings back in between $'s
    segments = wrapped.split('$')
    segments[1::2] = tex
    return '$'.join(segments)

if __name__ == '__main__':
    main()

You may use annotate with which you can specify offsets.

http://matplotlib.sourceforge.net/api/pyplot_api.html?highlight=annotate#matplotlib.pyplot.annotate

Or, you may consider to use the path_effect (available w/ v1.0).

http://matplotlib.sourceforge.net/examples/pylab_examples/patheffect_demo.html

Regards,

-JJ


On Wed, Nov 3, 2010 at 10:44 AM, Nikolaus Rath <Nik...@ra...> wrote:
> Hello,
>
> I am placing a text object in a place where the background is sometimes
> dark and sometimes light, so the text is hard to see. Adding a
> background to the text object itself looks a bit ugly, so I am wondering
> if there is a way to add a shadow to the text itself.
>
> What's the best way to do this?
>
> I thought about redrawing the same text just offset by a fraction of the
> text size, but I can't quite figure out how to do the placement
> properly. My rough idea is the get the position of the text in pixel
> coordinates, add an offset to it and then position it with display
> coordinates as well. But how do I get the display coordinates of
> something that I placed in axes coordinates? And how to I determine how
> many pixels I need to offset without knowing the dpi?
>
>
> Best,
>
>   -Nikolaus
>
> --
>  »Time flies like an arrow, fruit flies like a Banana.«
>
>  PGP fingerprint: 5B93 61F8 4EA2 E279 ABF6  02CF A9AD B7F8 AE4E 425C
>
> ------------------------------------------------------------------------------
> Achieve Improved Network Security with IP and DNS Reputation.
> Defend against bad network traffic, including botnets, malware,
> phishing sites, and compromised hosts - saving your company time,
> money, and embarrassment.   Learn More!
> http://p.sf.net/sfu/hpdev2dev-nov
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>

Hello,

I am placing a text object in a place where the background is sometimes
dark and sometimes light, so the text is hard to see. Adding a
background to the text object itself looks a bit ugly, so I am wondering
if there is a way to add a shadow to the text itself.

What's the best way to do this?

I thought about redrawing the same text just offset by a fraction of the
text size, but I can't quite figure out how to do the placement
properly. My rough idea is the get the position of the text in pixel
coordinates, add an offset to it and then position it with display
coordinates as well. But how do I get the display coordinates of
something that I placed in axes coordinates? And how to I determine how
many pixels I need to offset without knowing the dpi?


Best,

   -Nikolaus

-- 
 »Time flies like an arrow, fruit flies like a Banana.«

  PGP fingerprint: 5B93 61F8 4EA2 E279 ABF6  02CF A9AD B7F8 AE4E 425C

"Stan West" <sta...@pu...> writes:
>> From: Nikolaus Rath [mailto:Nik...@pu...] 
>> Sent: Saturday, October 30, 2010 16:17
>> 
>> but if I try to do the same think with the Y axis, everything looks
>> messed up (e.g. the matrix is no longer square):
>
> The matrix remains square for me using a build from Subversion.  What's your
> matplotlib.__version__?

In [16]: matplotlib.__version__
Out[16]: '1.0.0'

I attached the result of fig.savefig(). Let's see if it makes it through
the list.


Best,

   -Nikolaus

-- 
 »Time flies like an arrow, fruit flies like a Banana.«

  PGP fingerprint: 5B93 61F8 4EA2 E279 ABF6  02CF A9AD B7F8 AE4E 425C

A quick (and not safe) way w/ mpl v1.0 is,

    ax = plt.subplot(111)
    ax.plot(np.arange(3))

    ax.set_xticks([0, 0.5, 1., 1.5, 2.])

    mytick = ax.xaxis.majorTicks[2]
    mytick._apply_params(tickdir="out")

I don't think there is a way to do this only using public apis.
I myself actually prefer to create a separate axis (or axes), but this
will be more trickier than above solution.

Regards,

-JJ




On Tue, Nov 2, 2010 at 4:22 AM, Bartosz Telenczuk
<b.t...@bi...> wrote:
> Hi all,
>
> Is it possible to set direction (in or out) individually for each tick. I know about the rc setting ("(x/y)tick.direction") , but I need a finer control over the ticks.
>
> Thanks,
>
> Bartek
>
> ------------------------------------------------------------------------------
> Nokia and AT&T present the 2010 Calling All Innovators-North America contest
> Create new apps & games for the Nokia N8 for consumers in  U.S. and Canada
> $10 million total in prizes - $4M cash, 500 devices, nearly $6M in marketing
> Develop with Nokia Qt SDK, Web Runtime, or Java and Publish to Ovi Store
> http://p.sf.net/sfu/nokia-dev2dev
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>

I guess sometimes its best to dig into ipython. I apparently
misunderstood interp, but it seems it doesn't need to be used only
with grids created from a Basemap instance. I created the following
function which accomplishes my goals (I know this shouldn't all be in
one function, I'll look into making it more 'general' as a next step).
If I get around to cleaning it up, I'll repost.  --john

def create_land_veg_mask():
    """ Process to convert Gaussian grid data to lat lon (0.5 deg)
         for gridded data inside the grib2 ncep files available here:

         ftp://ftp.ncep.noaa.gov/pub/data/nccf/com/gfs/prod/
    """

    from mpl_toolkits.basemap import interp, shiftgrid
    import pygrib

    def interp_shift(data,lslons,lslats,lonsg,latsg,lons):
        dataout = interp(data.values,lslons[0,:],lslats[:,0],lonsg,latsg)
        data_shft,lon_shft = shiftgrid(-180,dataout.T,lons)
        return lon_shft, lats, data_shft

    grbs = pygrib.open('gfs.t00z.sfluxgrbf03.grib2')
    lsmask = grbs.select(name='Land-sea mask')[0]
    vegtypes = grbs.select(name='Vegetation Type')[0]
    lslats,lslons = lsmask.latlons() #same for vegtypes
    reg_lons = np.arange(-360,0,0.5)
    reg_lats = np.arange(-90,90,0.5)
    latsg,lonsg = np.meshgrid(reg_lats,reg_lons)

    lonsout, lats, vegout =
interp_shift(vegtypes,lslons,lslats,lonsg,latsg,reg_lons)
    lonsout, lats, lsmaskout =
interp_shift(lsmask,lslons,lslats,lonsg,latsg,lons)

    return lonsout,lats,vegout,lsmaskout

On 11/02/2010 04:39 AM, Alan G Isaac wrote:
> On 11/2/2010 10:06 AM, Benjamin Root wrote:
>> I have personally seen significant progress in this area, but there are a few backends that aren't quite right (MacOSX backend, I believe?).
>
>
> I believe multiple uses of ``show`` fail and are expected to fail
> for the foreseeable future when using IDLE to run a script.

This may be true in general; the effort has gone, and will continue to 
go, into getting consistent behavior in ipython and in standalone 
scripts, not in idle.  Nevertheless, the simple test running the 
attached script with idle on my linux machine (ubuntu 10.10), with mpl 
from svn, succeeded.  This might depend on the default backend, gtkagg 
in my case.

Eric


>
> fwiw,
> Alan Isaac

Hello, I have a set of data on a Gaussian grid. I have a latitude
array, a longitude array, and a data array:
In [121]: vlats.shape
Out[121]: (880, 1760)
In [122]: vlons.shape
Out[122]: (880, 1760)
In [123]: vtypes.shape
Out[123]: (880, 1760)

The longitude seems to go from -360 to 0.6, the latitude is more regular:
In [124]: vlons.min(); vlons.max()
Out[124]: -360.0
Out[124]: 0.59499999999880615

In [125]: vlats.min(); vlats.max()
Out[125]: -89.843513517868473
Out[125]: 89.843513517868473

I thought (and it was also suggested) I could use 'interp' from the
mpl_toolkits.basemap package, but I'm not sure how to do this. Does
anyone have experience with this? An alternative suggestion was to use
the scipy interp capabilities; again, I'm not certain I know how to go
about this.

Thanks,
john

I have a file now, that I created from a NCEP grib file. I'm trying to
get the land sea mask and the veg types into 0.5x0.5 lat/lon regular
grids. However, they are still in Gaussian Grids, but at least I have
it assembled into netcdf files! I'll follow up on a post about the
conversion. but I would be interested in knowing if what I have done
below fits the 'Convention 1.4' definitions:

In [64]: vegtype = grbs.select(name='Vegetation Type')[0]
In [65]: vlats,vlons = vegtype.latlons()
In [67]: vtypes = vegtype.values
In [70]: vtypes.shape
Out[70]: (880, 1760)
In [89]: nco = NetCDFFile('/wrk/jfb/DATASETS_TEMPLATES/JFB_vegtypes.nc','w',format='NETCDF4',zlib=True)
In [91]: nco.createDimension('nrows', 880)
Out[91]: <netCDF4.Dimension object at 0x8365550>
In [92]: nco.createDimension('ncols',1760)
Out[92]: <netCDF4.Dimension object at 0x8365b50>
In [93]: lat = nco.createVariable('lat','f8',('nrows','ncols'),zlib=True)
In [94]: lat[:] = vlats
In [95]: lon = nco.createVariable('lon','f8',('nrows','ncols'),zlib=True)
In [96]: lon[:] = vlons
In [97]: data = nco.createVariable('data','int',('nrows','ncols'),zlib=True)
In [98]: data[:] = vtypes
In [99]: nco.history="Creating a vegetation types mask from the NCEP
ncep.sfluxgrbf03.grb file"
In [100]: nco.units="Integer(0-13)"
In [101]: nco.close()
In [102]: nco =
NetCDFFile('/wrk/jfb/DATASETS_TEMPLATES/JFB_vegtypes.nc',format='NETCDF4',zlib=True)
In [103]: nco.variables.keys()
Out[103]: ['lat', 'lon', 'data']
In [104]: nco.close()

I guess officially, I should add something along the lines of:

float lat(lat) ;
 lat:long_name = "latitude" ;
 lat:units = "degrees_north" ;
 lat:standard_name = "latitude" ;

Thank you,
john



-- 
Configuration
``````````````````````````
Plone 2.5.3-final,
CMF-1.6.4,
Zope (Zope 2.9.7-final, python 2.4.4, linux2),
Python 2.6
PIL 1.1.6
Mailman 2.1.9
Postfix 2.4.5
Procmail v3.22 2001/09/10
Basemap: 1.0
Matplotlib: 1.0.0

Visit our website at http://www.ubs.com 

This message contains confidential information and is intended only 
for the individual named. If you are not the named addressee you 
should not disseminate, distribute or copy this e-mail. Please 
notify the sender immediately by e-mail if you have received this 
e-mail by mistake and delete this e-mail from your system. 

E-mails are not encrypted and cannot be guaranteed to be secure or 
error-free as information could be intercepted, corrupted, lost, 
destroyed, arrive late or incomplete, or contain viruses. The sender 
therefore does not accept liability for any errors or omissions in the 
contents of this message which arise as a result of e-mail transmission. 
If verification is required please request a hard-copy version. This 
message is provided for informational purposes and should not be 
construed as a solicitation or offer to buy or sell any securities 
or related financial instruments. 

UBS Limited is a company limited by shares incorporated in the United 
Kingdom registered in England and Wales with number 2035362. 
Registered office: 1 Finsbury Avenue, London EC2M 2PP.  UBS Limited 
is authorised and regulated by the Financial Services Authority. 

UBS AG is a public company incorporated with limited liability in 
Switzerland domiciled in the Canton of Basel-City and the Canton of 
Zurich respectively registered at the Commercial Registry offices in 
those Cantons with Identification No: CH-270.3.004.646-4 and having 
respective head offices at Aeschenvorstadt 1, 4051 Basel and 
Bahnhofstrasse 45, 8001 Zurich, Switzerland.  Registered in the 
United Kingdom as a foreign company with No: FC021146 and having a 
UK Establishment registered at Companies House, Cardiff, with No:  
BR 004507.  The principal office of UK Establishment: 1 Finsbury Avenue, 
London EC2M 2PP.  In the United Kingdom, UBS AG is authorised and 
regulated by the Financial Services Authority.

UBS reserves the right to retain all messages. Messages are protected 
and accessed only in legally justified cases. 

> From: Nikolaus Rath [mailto:Nik...@ra...] 
> Sent: Saturday, October 30, 2010 16:17
> 
> but if I try to do the same think with the Y axis, everything looks
> messed up (e.g. the matrix is no longer square):

The matrix remains square for me using a build from Subversion.  What's your
matplotlib.__version__?

From: Gino Serpa [mailto:gin...@gm...] 
Sent: Sunday, October 31, 2010 13:52
 
I am trying to change the format of the tick labels in a loglog plot from
exponential to decimal.
Right now the labels are   10^2 10 1 10^-1 10^-2
and I would like them to be   100 10 1 0.1 0.01

I tried finding easy explanations in the web with no luck. By the way I have
no clue what a class is and the word instantiation makes me crossed eyed

I hope you don't mind instantiating a class. See the documentation for tick
formatters [1], particularly FormatStrFormatter, and the major_minor_demo1
example [2].

[1] http://matplotlib.sourceforge.net/api/ticker_api.html#tick-formatting

[2]
http://matplotlib.sourceforge.net/examples/pylab_examples/major_minor_demo1.ht
ml

Hi

Thank you.
I installed the latest from svn and it now works.

Regards

Pål

On 2 November 2010 15:21, Benjamin Root <ben...@ou...> wrote:

> On Tue, Nov 2, 2010 at 8:20 AM, Pål Gunnar Ellingsen <pa...@gm...>wrote:
>
>> Hi
>>
>> I'm trying to use the linthershy option for symlog  in
>> matplotlib.axes.Axes.set_yscale() <http://matplotlib.sourceforge.net/api/axes_api.html#matplotlib.axes.Axes.set_yscale>for
>> values smaller than 1, but the plot then shows large values (vertical lines)
>> for the points which should have been in the linear range. My code is:
>>
>> import matplotlib.pyplot as plt
>> import numpy as np
>>
>>
>> f1 = np.loadtxt('someDataFile.txt') # file has all y values smaller than 1
>>
>> fig1 = plt.figure(1)
>> ax1 = fig1.add_subplot(111)
>>
>> ax1.plot(f1[:, 0], f1[:, 1])
>>
>> ax1.set_yscale('symlog', linthreshy = 1e-3)
>> ax1.set_ylim(1e-4, 1)
>> ax1.autoscale(enable = True, axis = 'x', tight = True)
>>
>> plt.draw()
>> plt.show()
>> ---------------
>> Another and testable code is
>>
>> import matplotlib
>> matplotlib.use('Qt4Agg')
>> import matplotlib.pyplot as plt
>> import numpy as np
>>
>> fig1 = plt.figure(1)
>> A = 10 + 9.999999999 * np.sin(2 * np.pi * np.arange(0.01, 1, 0.01))
>>
>> ax1 = fig1.add_subplot(111)
>> ax1.plot(A)
>>
>> ax1.set_yscale('symlog', linthreshy = 1e-6)
>> ax1.set_ylim(0, 20) # change this to ax1.set_ylim(0, 0.7) for strange
>> lines
>>
>>
>> fig2 = plt.figure(2)
>> ax2 = fig2.add_subplot(111)
>> ax2.plot(A)
>> ax2.set_yscale('log')
>> ax2.set_ylim(0, 20)
>> ax2.autoscale(enable = True, axis = 'both', tight = True)
>>
>> plt.draw()
>> plt.show()
>>
>> Which does not show correct labels and the linear part is gone (only a
>> straight line).
>> Does anyone have a solution for these problems?
>>
>> Regards
>>
>> Pål
>>
>>
> This was a bug found fairly recently.  In the case where the linthresh was
> less than 1.0:
>
>
> http://sourceforge.net/tracker/index.php?func=detail&aid=3081451&group_id=80706&atid=560720
>
> I don't know if the patch was backported to the maintenance branch, but I
> am fairly sure it would have been.
>
> I hope this was helpful,
> Ben Root
>
>

On 11/2/2010 10:06 AM, Benjamin Root wrote:
> I have personally seen significant progress in this area, but there are a few backends that aren't quite right (MacOSX backend, I believe?).


I believe multiple uses of ``show`` fail and are expected to fail
for the foreseeable future when using IDLE to run a script.

fwiw,
Alan Isaac

On Tue, Nov 2, 2010 at 8:20 AM, Pål Gunnar Ellingsen <pa...@gm...>wrote:

> Hi
>
> I'm trying to use the linthershy option for symlog  in
> matplotlib.axes.Axes.set_yscale() <http://matplotlib.sourceforge.net/api/axes_api.html#matplotlib.axes.Axes.set_yscale>for
> values smaller than 1, but the plot then shows large values (vertical lines)
> for the points which should have been in the linear range. My code is:
>
> import matplotlib.pyplot as plt
> import numpy as np
>
>
> f1 = np.loadtxt('someDataFile.txt') # file has all y values smaller than 1
>
> fig1 = plt.figure(1)
> ax1 = fig1.add_subplot(111)
>
> ax1.plot(f1[:, 0], f1[:, 1])
>
> ax1.set_yscale('symlog', linthreshy = 1e-3)
> ax1.set_ylim(1e-4, 1)
> ax1.autoscale(enable = True, axis = 'x', tight = True)
>
> plt.draw()
> plt.show()
> ---------------
> Another and testable code is
>
> import matplotlib
> matplotlib.use('Qt4Agg')
> import matplotlib.pyplot as plt
> import numpy as np
>
> fig1 = plt.figure(1)
> A = 10 + 9.999999999 * np.sin(2 * np.pi * np.arange(0.01, 1, 0.01))
>
> ax1 = fig1.add_subplot(111)
> ax1.plot(A)
>
> ax1.set_yscale('symlog', linthreshy = 1e-6)
> ax1.set_ylim(0, 20) # change this to ax1.set_ylim(0, 0.7) for strange lines
>
>
> fig2 = plt.figure(2)
> ax2 = fig2.add_subplot(111)
> ax2.plot(A)
> ax2.set_yscale('log')
> ax2.set_ylim(0, 20)
> ax2.autoscale(enable = True, axis = 'both', tight = True)
>
> plt.draw()
> plt.show()
>
> Which does not show correct labels and the linear part is gone (only a
> straight line).
> Does anyone have a solution for these problems?
>
> Regards
>
> Pål
>
>
This was a bug found fairly recently.  In the case where the linthresh was
less than 1.0:

http://sourceforge.net/tracker/index.php?func=detail&aid=3081451&group_id=80706&atid=560720

I don't know if the patch was backported to the maintenance branch, but I am
fairly sure it would have been.

I hope this was helpful,
Ben Root

On Tue, Nov 2, 2010 at 9:00 AM, John <was...@gm...> wrote:

> Thanks Ben!
>
> I guess where I am with this now is that I have found, as you say, the
> VGTYP in the
> operational products (the *sflux* grib files), but I don't find it on
> any of the 0.5 x 0.5 degree files. Which means, that as you say, I'll
> have to figure out how to reproject the data from the Gaussian grid to
> a regular grid with 0.5x0.5 lat lon coordinates.
>
> Any thoughts on that process?? :D
>
> Thanks again,
> john
>
>
>
I haven't tried doing conversions between two different projections, but
maybe the Basemap package might be useful?

http://matplotlib.sourceforge.net/basemap/doc/html/

Basically, you could get the arrays of x and y positions of your data, and
then determine what would be the x and y positions of the projection that
you want, and then use scipy's griddata to produce the results on the grid
you want:

http://docs.scipy.org/doc/scipy/reference/generated/scipy.interpolate.griddata.html#scipy.interpolate.griddata

This is a bit brute-force, and I am sure there are more elegant solutions,
but that should work in a pinch.

Ben Root

On Tue, Nov 2, 2010 at 8:31 AM, Daπid <dav...@gm...> wrote:

> On Tue, Nov 2, 2010 at 12:49 PM, David Kremer <dav...@gm...>
> wrote:
> > Personally I used show() yesterday, and it blocks the execution until
> > the figure's window is closed. It is of perfect convenience I think,
> > as a default behavior.
>
> Yes, this is the optimal behavior, but it is not fully implemented, as
> seen in the documentation.
>
>
I have personally seen significant progress in this area, but there are a
few backends that aren't quite right (MacOSX backend, I believe?).  Anyway,
some effort was made right before the 1.0.0 release, but it was soon
discovered that it was still not right (and even caused some regressions, I
believe).  So, shortly there-after, a lot of work went into making the
behavior more uniform across all backends.

If the show() behavior is currently not working properly for you, I would
suggest building from source (if possible) the maintenance v1.0.0 branch
which has it mostly right (I believe the only exception is the macosx
backend).  If that isn't possible, there have been talks of soon releasing a
v1.0.1 package which would contain all of those fixes.

Note, I am not an authority in this matter as I was only just joining the
development group at around that time, but I hope this answers your
question.

Ben Root

Thanks Ben!

I guess where I am with this now is that I have found, as you say, the
VGTYP in the
operational products (the *sflux* grib files), but I don't find it on
any of the 0.5 x 0.5 degree files. Which means, that as you say, I'll
have to figure out how to reproject the data from the Gaussian grid to
a regular grid with 0.5x0.5 lat lon coordinates.

Any thoughts on that process?? :D

Thanks again,
john





>
> John,
>
> GRIB files are an interesting beast.  It has been awhile since I have dealt
> with them, but I will try to help you with what I can recall.  First, the
> data records in the files are indicated by:
>
>    1. Record number
>    2. Position in bytes
>    3. Date (YYMMDDHH).
>    4. Parameter name (LAND=land/sea mask)
>    5. Indicator of parameter and units (grib PDS octet 9)
>    6. Type of level/layer (grib PDS octet 10)
>
>
>    7. Height, pressure, etc (grib PDS octets 11-12)
>    8. Time Range (grib PDS octet 21)
>    9. Period of time 1, (grib PDS octet 19)
>   10. Period of time 2, (grib PDS octet 20)
>   11. Forecast time unit (grib PDS octet 18)
>
>
>   12. level
>   13. anl=analysis, fcst=forecast
>   14. NAve (number of grids used to make average)
>
> The "type of level/layer" I believe is the grid projection type.  Anyway,
> these things can be mixed and matched, so all we need to do is find your
> parameter in the grid projection that you want (or you will have to
> re-project the data).
>
> The operational models appear to have VGTYP, but the reanalysis products do
> not appear to have this.  Maybe this is causing your confusion?
>
> This page might be more useful to you:
> http://www.cpc.noaa.gov/products/wesley/wgrib.html
>
> I hope this is useful,
> Ben Root
>
>



-- 
Configuration
``````````````````````````
Plone 2.5.3-final,
CMF-1.6.4,
Zope (Zope 2.9.7-final, python 2.4.4, linux2),
Python 2.6
PIL 1.1.6
Mailman 2.1.9
Postfix 2.4.5
Procmail v3.22 2001/09/10
Basemap: 1.0
Matplotlib: 1.0.0

On Tue, Nov 2, 2010 at 5:19 AM, John <was...@gm...> wrote:

> Maybe I should also follow up on this and indicate that one variable I
> really want is the Veg Type defiinitions from the files. It exists in
> the file from joewheatley's example, (VGTYP_P0_L1_GGA0) but it is on a
> Gaussian lat/lon grid. I've downloaded some of the gfs files from the
> 0.5 x 0.5 degree output, but I don't find the variable.
>
> Thanks again!
>
> On Tue, Nov 2, 2010 at 10:33 AM, John <was...@gm...> wrote:
> > Folks,
> >
> > I'm trying to find documentation online that provides guidance on what
> > all the variables are that are in the NCEP model output grib files.
> > There are many NCEP models, so to be clear, I'm referring to the GFS
> > 0.5x0.5 degree product. I was looking at this nice blog:
> >
> > http://joewheatley.net/ncep-global-forecast-system/
> >
> > and the files that are available here:
> >  ftp://ftp.ncep.noaa.gov/pub/data/nccf/com/gfs/prod
> >
> > And I see this link:
> > http://www.nco.ncep.noaa.gov/pmb/products/gfs/
> >
> > But looking at a file and the variables:
> > In [112]: ncep.variables.keys()
> > Out[112]:
> > ['PRES_P8_L223_GLL0_avg',
> >  'TOZNE_P0_L200_GLL0',
> >  'PRES_P0_L242_GLL0',
> >  'HGT_P0_L100_GLL0',
> >  'VVEL_P0_L100_GLL0',
> >  'UGRD_P0_L6_GLL0',
> >  'HGT_P0_L7_GLL0',
> >  'CIN_P0_2L108_GLL0',
> >  'ULWRF_P8_L8_GLL0_avg',
> >  'PRES_P0_L109_GLL0',
> > (snip)
> >
> > is somewhat clear in most cases, but is there a table which directly
> > maps these online?. Also, you can of course look at the 'long_name',
> > 'level', and 'level_type' attributes of each variable, but I'm looking
> > more or less for an online table showing all this. Maybe I need to
> > script it quickly ;)
> >
> > Thanks,
> > john
> >
>
>
John,

GRIB files are an interesting beast.  It has been awhile since I have dealt
with them, but I will try to help you with what I can recall.  First, the
data records in the files are indicated by:

   1. Record number
   2. Position in bytes
   3. Date (YYMMDDHH).
   4. Parameter name (LAND=land/sea mask)
   5. Indicator of parameter and units (grib PDS octet 9)
   6. Type of level/layer (grib PDS octet 10)

   7. Height, pressure, etc (grib PDS octets 11-12)
   8. Time Range (grib PDS octet 21)
   9. Period of time 1, (grib PDS octet 19)
  10. Period of time 2, (grib PDS octet 20)
  11. Forecast time unit (grib PDS octet 18)

  12. level
  13. anl=analysis, fcst=forecast
  14. NAve (number of grids used to make average)

The "type of level/layer" I believe is the grid projection type.  Anyway,
these things can be mixed and matched, so all we need to do is find your
parameter in the grid projection that you want (or you will have to
re-project the data).

The operational models appear to have VGTYP, but the reanalysis products do
not appear to have this.  Maybe this is causing your confusion?

This page might be more useful to you:
http://www.cpc.noaa.gov/products/wesley/wgrib.html

I hope this is useful,
Ben Root

On Tue, Nov 2, 2010 at 12:49 PM, David Kremer <dav...@gm...> wrote:
> Personally I used show() yesterday, and it blocks the execution until
> the figure's window is closed. It is of perfect convenience I think,
> as a default behavior.

Yes, this is the optimal behavior, but it is not fully implemented, as
seen in the documentation.

Hi

I'm trying to use the linthershy option for symlog  in
matplotlib.axes.Axes.set_yscale()
<http://matplotlib.sourceforge.net/api/axes_api.html#matplotlib.axes.Axes.set_yscale>for
values smaller than 1, but the plot then shows large values (vertical lines)
for the points which should have been in the linear range. My code is:

import matplotlib.pyplot as plt
import numpy as np


f1 = np.loadtxt('someDataFile.txt') # file has all y values smaller than 1

fig1 = plt.figure(1)
ax1 = fig1.add_subplot(111)

ax1.plot(f1[:, 0], f1[:, 1])

ax1.set_yscale('symlog', linthreshy = 1e-3)
ax1.set_ylim(1e-4, 1)
ax1.autoscale(enable = True, axis = 'x', tight = True)

plt.draw()
plt.show()
---------------
Another and testable code is

import matplotlib
matplotlib.use('Qt4Agg')
import matplotlib.pyplot as plt
import numpy as np

fig1 = plt.figure(1)
A = 10 + 9.999999999 * np.sin(2 * np.pi * np.arange(0.01, 1, 0.01))

ax1 = fig1.add_subplot(111)
ax1.plot(A)

ax1.set_yscale('symlog', linthreshy = 1e-6)
ax1.set_ylim(0, 20) # change this to ax1.set_ylim(0, 0.7) for strange lines


fig2 = plt.figure(2)
ax2 = fig2.add_subplot(111)
ax2.plot(A)
ax2.set_yscale('log')
ax2.set_ylim(0, 20)
ax2.autoscale(enable = True, axis = 'both', tight = True)

plt.draw()
plt.show()

Which does not show correct labels and the linear part is gone (only a
straight line).
Does anyone have a solution for these problems?

Regards

Pål