matplotlib-users Mailing List for matplotlib (Page 4)

Ken McIvor wrote:
> On Jan 21, 2007, at 11:04 AM, Allan Noriel Estrella wrote:
>> does matplotlib already support wxPython 2.8?
> 
> I have not had the opportunity to test it with 2.8. However, I am 
> given to understand that it works. The one caveat is that you cannot 
> presently compile the WXAgg accelerator with 2.8.
I've tested it with 2.8 on OS-X, and it can be compiled with an almost 
trivial fix.
There is an odd bug with the standard pylab toolbar: the icons on the 
buttons disappear after they have been clicked once, though the buttons 
do still work.
What platform do you want it on? Are you set up to compile it yourself?
NOTE: wxPython2.8 has added some features for more directly creating 
bitmaps from data. It would be great if someone used those in the 
back-end -- if it works well, we might be able to get rid of the 
accelerated back end altogether. You can search this list for more info 
-- look for posts by me and Ken McIvor.
Here's a bit more detail about the current state of affairs (this was 
recently posted on matplotlib-devel:
Russell Owen just built an installer for MPL on OS_X for Python2.5,
wxPython2.8.
To do it, he needed to patch
_wxagg.cpp, at line 238 as follows:
OLD:
wxBitmap *bitmap = new wxBitmap(image);
NEW
wxBitmap *bitmap = new wxBitmap(*image);
Thanks to a hint from Robin Dunn.
It now seems to work OK, except that when you use pylab.show() and then
click on the toolbar buttons to zoom, etc, the buttons no longer display
Also, it doesn't work with Numeric 24.2 either -- I think that's a known
issue, but I can't find a note about it at the moment.
Is anyone maintaining the wx back-end now?
Is MPL working with wxPython 2.8 on other platforms?
If anyone wants to try this binary out, it's temporarily at:
<http://www.astro.washington.edu/rowen/pythonpackages/Python%202.5/>
 >>> matplotlib.__version__
'0.87.7'
 >>> wx.__version__
'2.8.0.1'
 >>> numpy.__version__
'1.0.1'
 >>> Numeric.__version__
'24.2'
-Chris
-- 
Christopher Barker, Ph.D.
Oceanographer
Emergency Response Division
NOAA/NOS/OR&R (206) 526-6959 voice
7600 Sand Point Way NE (206) 526-6329 fax
Seattle, WA 98115 (206) 526-6317 main reception
Chr...@no...

Hi,
is there a way to make a bar plot with two different y axes in matplotlib?
Something similar to plotyy in matlab but for a bar plot?
Thanks and regards
Francesco

I would like to be able to draw rectangles, with text centered inside
them, at key positions on a plot.
 
The case in question is a bar plot, and the rectangles need to be drawn
such that they are length of one or more of the each bar groupings,
side-by-side in a line along the width of the X-axis. The line of
rectangles should be located just below the X-axis tickmark labels.
 
I see the fill() command allows plotting of a polygon and the figtext()
command allows text to be plotted so, in theory, these should work.
 
My problem is 
 
(a) what is the syntax for the fill() command - I have tried
ax.fill(0.5,0.5,'b') and get an error:
 TypeError: zip argument #1 must support iteration
 
(b) how to calculate the correct (x,y) values for locating the
start/stop points of the rectangles? i.e. how do I find out where the
start/end points are for the X values of each bar grouping? And at what
vertical height should the line of rectangles be located?
 
Any ideas or insights would be appreciated?
 
Thanks
Derek
-- 
This message is subject to the CSIR's copyright, terms and conditions and
e-mail legal notice. Views expressed herein do not necessarily represent the
views of the CSIR.
 
CSIR E-mail Legal Notice
http://mail.csir.co.za/CSIR_eMail_Legal_Notice.html 
 
CSIR Copyright, Terms and Conditions
http://mail.csir.co.za/CSIR_Copyright.html 
 
For electronic copies of the CSIR Copyright, Terms and Conditions and the CSIR
Legal Notice send a blank message with REQUEST LEGAL in the subject line to
Cal...@cs....
This message has been scanned for viruses and dangerous content by MailScanner, 
and is believed to be clean.

On Jan 21, 2007, at 11:04 AM, Allan Noriel Estrella wrote:
>
> does matplotlib already support wxPython 2.8?
I have not had the opportunity to test it with 2.8. However, I am 
given to understand that it works. The one caveat is that you cannot 
presently compile the WXAgg accelerator with 2.8. Everything will 
still work without, but animation is slower.
Ken

does matplotlib already support wxPython 2.8?

>>>>> "Francesco" == Francesco Davide Calabrese <fd....@gm...> writes:
 Francesco> Hi, the following simple code (in Matlab syntax):
 Francesco> x1 = [5, 6] y1 = [3, 2] bar([1,2], [x1; y1])
 Francesco> produces a bar plot where two of the bars have a color
 Francesco> (let's say red) and the two other bars have another
 Francesco> color (let's say blue). I'd like to know whether it's
 Francesco> possible in matplotlib to plot the bars not simply with
 Francesco> a different color but with a different filling
 Francesco> pattern. So, for example, the blue bars filled with
 Francesco> vertical lines (or something else) and the red bars
 Francesco> filled with horizontal lines (or again something else).
 Francesco> Please, note that I'm not talking about the borders of
 Francesco> the bars. The reason why I need this feature is that I
 Francesco> have to clearly distinguish the bars also when they are
 Francesco> plotted in black and white, because many publications
 Francesco> can't be done in colors.
The postscript backend supports this -- we call it "hatching".
Unfortunately, the others do not yet. See the example at
http://matplotlib.sf.net/examples/hatch_demo.py
One could be clever with "imshow" and create arrays that have the
hatching. In the example I wrote below I created a custom bar chart
using gradient images. By adding horizontal, vertical, or diangonal
lines to the arrays, or dots, one could support various hatching
schemes. Here is the sample code for gradients that i wrote in
response to an earlier post:
 from pylab import figure, show, nx, cm
 def gbar(ax, x, y, width=0.5, bottom=0):
 X = [[.6, .6],[.7,.7]]
 for left,top in zip(x, y):
 right = left+width
 ax.imshow(X, interpolation='bicubic', cmap=cm.Blues,
 extent=(left, right, bottom, top), alpha=1)
 fig = figure()
 xmin, xmax = xlim = 0,10
 ymin, ymax = ylim = 0,1
 ax = fig.add_subplot(111, xlim=xlim, ylim=ylim,
 autoscale_on=False)
 X = [[.6, .6],[.7,.7]]
 ax.imshow(X, interpolation='bicubic', cmap=cm.copper,
 extent=(xmin, xmax, ymin, ymax), alpha=1)
 N = 10
 x = nx.arange(N)+0.25
 y = nx.mlab.rand(N)
 gbar(ax, x, y, width=0.7)
 ax.set_aspect('normal')
 show()
Ideally, we could do the work in Agg to support these hatched
images....
JDH

On Jan 21, 2007, at 8:21 AM, Allan Noriel Estrella wrote:
>
> i found this code snippet in one example in matplotlib
>
<snip>
>
> i just want to ask what does app = App(0) mean, why do we need 0 
> for an argument
It explicitly disables the redirection of sys.stdout and sys.stderr.
	http://www.wxpython.org/docs/api/wx.App-class.html#__init__
Ken

i found this code snippet in one example in matplotlib
class App(wx.App):
 def OnInit(self):
 'Create the main window and insert the custom frame'
 frame = CanvasFrame()
 frame.Show(True)
 return True
app = App(0)
app.MainLoop()
i just want to ask what does app = App(0) mean, why do we need 0 for an
argument

Hi,
the same problem occurred, when I also tried to plot some chaos systems
the solution I found (and works just fine) is to comment axes.py:2131
I didn't send a patch, 'cos I'm not sure that it wasn't really an
common bug (I played with my overall python distribution)
some OT:
1) trajectories -> you could use pydoc numpy.hsplit
2) IMHO to plot chaotic behaviour with iterations over 500k the matplotli=
b on
an ordinary computer is a little bit slow compared to creating a data
file and using gnuplot ...
Karol
On Fri, Jan 12, 2007 at 11:17:30AM +0100, Mico Fil=F3s wrote:
> Hi all
>=20
> I am trying to plot the trajectory of the Lorenz system with the axes3d=
.py
> module (version 0.87.7). The code is the following:
>=20
...
>=20
> I get however some errors related with the autoscale_view() method (see=
 below).
> I have two questions: Is plot3D the correct choice for plotting a 3D
> curve? In that case, is there any way to fix these errors?
>=20
> Thanks a lot for your help and for making matplotlib possible,
>=20
> dani
>=20
> Errors (pdb output):
>=20
> lorenz.py
> ---> 42 plot3D(x,y,z)
>=20
> /usr/lib/python2.4/site-packages/matplotlib/axes3d.py in plot3D(self,
> xs, ys, zs, *args, **kwargs)
> 488 def plot3D(self, xs, ys, zs, *args, **kwargs):
> 489 had_data =3D self.has_data()
> --> 490 lines =3D Axes.plot(self, xs,ys, *args, **kwargs)
> 491 if len(lines)=3D=3D1:
> 492 line =3D lines[0]
>=20
> /usr/lib/python2.4/site-packages/matplotlib/axes.py in plot(self,
> *args, **kwargs)
> 2129 lines =3D [line for line in lines] # consume the genera=
tor
> 2130
> -> 2131 self.autoscale_view(scalex=3Dscalex, scaley=3Dscaley)
> 2132 return lines
>=20
> TypeError: autoscale_view() got an unexpected keyword argument 'scalex'
> > /usr/lib/python2.4/site-packages/matplotlib/axes.py(2131)plot()
> 2130
> -> 2131 self.autoscale_view(scalex=3Dscalex, scaley=3Dscaley)
> 2132 return lines
>=20

Hi,
the following simple code (in Matlab syntax):
x1 = [5, 6]
y1 = [3, 2]
bar([1,2], [x1; y1])
produces a bar plot where two of the bars have a color (let's say red) and
the two other bars have another color (let's say blue). I'd like to know
whether it's possible in matplotlib to plot the bars not simply with a
different color but with a different filling pattern. So, for example, the
blue bars filled with vertical lines (or something else) and the red bars
filled with horizontal lines (or again something else).
Please, note that I'm not talking about the borders of the bars. The reason
why I need this feature is that I have to clearly distinguish the bars also
when they are plotted in black and white, because many publications can't be
done in colors.
Any help will be very appreciated.
Thanks and regards
Francesco

Hi Folks,
 
 I was wondering how to set the Matplotlib classic Toolbar in order to display only existent data when the back/forward arrows are clicked (i.e, the pan feature) for a subplot that displays plotted data. 
 
 How to set the code in the matplotlib toolbar to only display the data from the file? ( clicking the left arrow when the left most point is at the start of the data file should result in freezing of the plot and a similar action when the right arrow is clicked when the right most point is at the end of the data file). 
 
 I'm not sure if my post was sufficiently as clear as spring water. Any suggestions would be greatly appreciated.
 
 Great forum, by the way.. 
 
 Thanks.
 -Maser
 
---------------------------------
Food fight? Enjoy some healthy debate
in the Yahoo! Answers Food & Drink Q&A.

Tim,
On 1/19/07, tim...@gm... <tim...@gm...> wrote=
:
> how can I read data in that
> x gets the latitude
> y gets the longgitute
> z gets the data which will be shown in different color accoring to their =
values?
Use pylab.load to load the data into an array. Then, the rest of what
you want to accomplish is detailed in the wiki:
<http://www.scipy.org/Cookbook/Matplotlib/Maps>.
Hope that helps!
Jos=E9

if you're building an app i would suggest using the OO interface.
There are good examples in the examples directory. But basically you
get a figure either by constructing it or using pylab.gcf and you can
add_axes to the figure and plot on them. It is much more scalable for
an app than the pylab interface.
as for your original question, i'm afraid i've never seen the behavior
before and don't have a good answer.
--Tom
On 1/19/07, Jonathon Anderson <and...@gm...> wrote:
> It's definitely not the behavior I'm seeing here. In my matploblibrc file,
> hold is set to True. Is there another value that might be influencing this?
>
> In any case, I don't want the behavior to be dependent on a config file: I'm
> building an application. Do you know how I might specify this behavior at
> run-time?
>
> ~jonathon
>
>
> On 1/19/07, Tom Denniston
> <tom...@al...> wrote:
> > It might depend on what's in your matplotlib rc file but by default
> > the behavior I have always seen was each plot command adds to the
> > current figure (pylab.gcf()) until you do a pylab.clf().
> >
> > So the two commands lists you have below end up being functionally
> equivalent.
> >
> > --Tom
> >
> > On 1/19/07, Jonathon Anderson < and...@gm...> wrote:
> > > I have several lines of data that I want to plot on the same graph, but
> > > every time I run the pylab.plot() function it redraws the graph from
> > > nothing. I've tried pylab.plot(*, hold=True) and pylab.hold(True), but
> it
> > > still happens. Can I add data to an existing figure, or do I have to
> pass
> > > all the data at once?
> > >
> > > e.g., Do I have to do this:
> > >
> > > pylab.plot(x1, y1, x2, y2, x3, y3, ...)
> > >
> > > or can I do this:
> > >
> > > pylab.plot(x1, y1)
> > > pylab.plot(x2, y2)
> > > pylab.plot(x3, y3)
> > > ...
> > >
> > > ~jonathon anderson
> > >
> > >
> -------------------------------------------------------------------------
> > > Take Surveys. Earn Cash. Influence the Future of IT
> > > Join SourceForge.net's Techsay panel and you'll get the chance to share
> your
> > > opinions on IT & business topics through brief surveys - and earn cash
> > >
> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
> > >
> > > _______________________________________________
> > > Matplotlib-users mailing list
> > > Mat...@li...
> > >
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> > >
> > >
> > >
> >
>
>
> -------------------------------------------------------------------------
> Take Surveys. Earn Cash. Influence the Future of IT
> Join SourceForge.net's Techsay panel and you'll get the chance to share your
> opinions on IT & business topics through brief surveys - and earn cash
> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
>
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
>
>

It's definitely not the behavior I'm seeing here. In my matploblibrc file,
hold is set to True. Is there another value that might be influencing this?
In any case, I don't want the behavior to be dependent on a config file: I'm
building an application. Do you know how I might specify this behavior at
run-time?
~jonathon
On 1/19/07, Tom Denniston <tom...@al...> wrote:
>
> It might depend on what's in your matplotlib rc file but by default
> the behavior I have always seen was each plot command adds to the
> current figure (pylab.gcf()) until you do a pylab.clf().
>
> So the two commands lists you have below end up being functionally
> equivalent.
>
> --Tom
>
> On 1/19/07, Jonathon Anderson <and...@gm...> wrote:
> > I have several lines of data that I want to plot on the same graph, but
> > every time I run the pylab.plot() function it redraws the graph from
> > nothing. I've tried pylab.plot(*, hold=True) and pylab.hold(True), but
> it
> > still happens. Can I add data to an existing figure, or do I have to
> pass
> > all the data at once?
> >
> > e.g., Do I have to do this:
> >
> > pylab.plot(x1, y1, x2, y2, x3, y3, ...)
> >
> > or can I do this:
> >
> > pylab.plot(x1, y1)
> > pylab.plot(x2, y2)
> > pylab.plot(x3, y3)
> > ...
> >
> > ~jonathon anderson
> >
> >
> -------------------------------------------------------------------------
> > Take Surveys. Earn Cash. Influence the Future of IT
> > Join SourceForge.net's Techsay panel and you'll get the chance to share
> your
> > opinions on IT & business topics through brief surveys - and earn cash
> >
> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
> >
> > _______________________________________________
> > Matplotlib-users mailing list
> > Mat...@li...
> > https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> >
> >
> >
>

I have several lines of data that I want to plot on the same graph, but
every time I run the pylab.plot() function it redraws the graph from
nothing. I've tried pylab.plot(*, hold=True) and pylab.hold(True), but it
still happens. Can I add data to an existing figure, or do I have to pass
all the data at once?
e.g., Do I have to do this:
pylab.plot(x1, y1, x2, y2, x3, y3, ...)
or can I do this:
pylab.plot(x1, y1)
pylab.plot(x2, y2)
pylab.plot(x3, y3)
...
~jonathon anderson

>>>>> "Matthew" == Matthew Koichi Grimes <mk...@cs...> writes:
 Matthew> Autoscaling is usually cool, but for my particular
 Matthew> application I'd like to turn it off, so that my
 Matthew> successive surface plots are all shown in the same
 Matthew> scale. How can I turn off autoscaling in Axes3D?
 Matthew> My current workaround is to manually set the axis limits
 Matthew> just before each draw(), but this is kind of a pain, and
 Matthew> fragile.
Matt, just to let you know, axes3d is not really supported. We
include it because some people find it useful and we're hoping some
person (you!) will pick it up and run with it. This code was written
and generously donated by John Porter, who is too busy to maintain it
as are the other core mpl devels. So please use it and post problems,
but don't be surprised if noone comes to your aid. Most likely, you
will need to find and fix these bugs yourself as you can.
In general, you can turn off autoscaling with 
 ax.set_autoscale_on(False)
Though I would have to dig through the code to see if this is
supported for Axes3d....
JDH

Hi!
I am farely new to matplotlib and basemap. I have one question about how to plot data from a csv file on a map.
the header goes as this
ID lat(x) long(y) measurements(z)
how can I read data in that
x gets the latitude
y gets the longgitute
z gets the data which will be shown in different color accoring to their values?
I played around and started from quiver_demo.py but didn't find my way.
reader = csv.reader(open('t.csv'))
for row in reader:
 print row
Could you give me some hints or starting points?
Thanks,
Tim
-- 
Der GMX SmartSurfer hilft bis zu 70% Ihrer Onlinekosten zu sparen! 
Ideal für Modem und ISDN: http://www.gmx.net/de/go/smartsurfer

Autoscaling is usually cool, but for my particular application I'd like 
to turn it off, so that my successive surface plots are all shown in the 
same scale. How can I turn off autoscaling in Axes3D?
My current workaround is to manually set the axis limits just before 
each draw(), but this is kind of a pain, and fragile.
-- Matt

I posted a question earlier about how to repeatedly plot (i.e. animate) 
a changing surface without having multiple surfaces just accumulate in 
the plot. Eric Firing suggested using Axes3DI.hold(False), but this 
didn't work; is that a bug, or is that expected? If it's a bug, how can 
I file it?
I'm using matplotlib 0.87.7. To reproduce this weirdness, run the 
following code:
# some code to illustrate that Axes3D.hold() doesn't work as expected
import numpy as N
import pylab as P
import matplotlib.axes3d as P3
P.ion() 
fig = P.gcf()
ax3d = P3.Axes3D(fig)
xGrid, yGrid = P.meshgrid(*[N.linspace(-1., 1., 100)]*2)
zGrid = xGrid**2+yGrid**2
numFrames = 10
scales = N.linspace(-1., 1., numFrames)
for scale in scales:
 ax3d.plot_surface( xGrid, yGrid, zGrid*scale )
 ax3d.hold(True)
 ax3d.contour3D( xGrid, yGrid, zGrid*scale )
 ax3d.hold(False)
 P.draw()
# Note that using P.hold instead of ax3d.hold doesn't work either.
-- Matt

Hello,
i have made bar charts but I was wondering if there is a way to show on top of each bar
the real content/counts, like a label; I generate static plots so the dynamic showing of the
counts on focusing on the bar does not help me. 
Thank you much for some feedback.
best regards
Margherita

Hello list!
Is there a way to avoid the thin line at the end/beginning of data when
using an orthographic projection? With other projections or when no pole
is visible, the line can be hidden by shifting the data grid (as done in
the example), but if you have to have one of the poles visible, I don't
know how to get rid of the line.
Example (modified contour_demo.py from examples):
from matplotlib.toolkits.basemap import Basemap, shiftgrid
from pylab import *
# examples of filled contour plots on ortho projection.
# read in data on lat/lon grid.
hgt = load('500hgtdata.gz')
lons = load('500hgtlons.gz')
lats = load('500hgtlats.gz')
# shift data so lons go from -180 to 180 instead of 0 to 360.
hgt,lons = shiftgrid(180.,hgt,lons,start=False)
lons, lats = meshgrid(lons, lats)
# setup of orthographic basemap
m = Basemap(resolution='c',projection='ortho',\
 lat_0=50.,lon_0=180.)
fig=figure()
ax = fig.add_axes([0.1,0.1,0.7,0.7])
# make a filled contour plot.
x, y = m(lons, lats)
CS = m.contour(x,y,hgt,15,linewidths=0.5,colors='k')
CS = m.contourf(x,y,hgt,15,cmap=cm.jet)
l,b,w,h=ax.get_position()
cax = axes([l+w+0.075, b, 0.05, h]) # setup colorbar axes
colorbar(drawedges=True, cax=cax) # draw colorbar
axes(ax) # make the original axes current again
# draw coastlines and political boundaries.
m.drawcoastlines()
m.fillcontinents()
m.drawmapboundary()
# draw parallels and meridians.
#parallels = arange(-80.,90,20.)
#m.drawparallels(parallels)
#meridians = arange(0.,360.,20.)
#m.drawmeridians(meridians)
title('Orthographic Filled Contour Demo')
show()

>>>>> "Derek" == Derek Hohls <DH...@cs...> writes:
 Derek> The command: ax = lab.subplot(111, axisbelow='True')
 
 Derek> ensures that the gridlines appear below the bars on a plot;
 Derek> how can the same be done for the legend (ie the grid below
 Derek> the legend and not running through it)?
Hmm, I don't see this in matplotlib svn. Ie, in the example below, the
grid is below the legend.
 In [1]: ax = subplot(111, axisbelow=True)
 In [2]: plot([1,2,3], label='test')
 Out[2]: [<matplotlib.lines.Line2D instance at 0xb581b6ac>]
 In [3]: legend()
 Out[3]: <matplotlib.legend.Legend instance at 0xb581bc4c>
 In [4]: grid()
 
If this doesn't work for you and you cannot upgrade, you may be able
to increase the zorder of your legend 
 l = legend(...)
 l.set_zorder(20)
and see if that helps.
JDH

>>>>> "Derek" == Derek Hohls <DH...@cs...> writes:
 Derek> legend( ... prop=FontProperties('smaller') )
 
 Derek> as per the webpage, I get a warning message and the font
 Derek> size is unchanged.
 
 Derek> What is the best method for changing font properties
 Derek> (e.g. point size, color) for the legend?
 
I'm not sure why this isn't working, but one approach is to get all
the text instances from the legend and set their properties directly
l = legend(...)
for t in l.get_texts():
 t.set_fontsize(12)
 t.set_color('red')
See also 
http://matplotlib.sourceforge.net/examples/legend_demo.py
JDH

I would like to know how to change the font for the legend.
 
The usual method of:
 
 yticklabels = getp(gca(), 'yticklabels')
 setp(yticklabels, color='black', fontsize=6)
 
Does not seem to work for legend. 
 
And when I try:
 
legend( ... prop=FontProperties('smaller') )
 
as per the webpage, I get a warning message and the 
font size is unchanged.
 
What is the best method for changing font properties
(e.g. point size, color) for the legend?
 
Thanks
Derek
-- 
This message is subject to the CSIR's copyright, terms and conditions and
e-mail legal notice. Views expressed herein do not necessarily represent the
views of the CSIR.
 
CSIR E-mail Legal Notice
http://mail.csir.co.za/CSIR_eMail_Legal_Notice.html 
 
CSIR Copyright, Terms and Conditions
http://mail.csir.co.za/CSIR_Copyright.html 
 
For electronic copies of the CSIR Copyright, Terms and Conditions and the CSIR
Legal Notice send a blank message with REQUEST LEGAL in the subject line to
Cal...@cs....
This message has been scanned for viruses and dangerous content by MailScanner, 
and is believed to be clean.

The command:
 
ax = lab.subplot(111, axisbelow='True') 
 
ensures that the gridlines appear below the bars on a plot;
how can the same be done for the legend (ie the grid
below the legend and not running through it)?
 
Thanks
Derek
-- 
This message is subject to the CSIR's copyright, terms and conditions and
e-mail legal notice. Views expressed herein do not necessarily represent the
views of the CSIR.
 
CSIR E-mail Legal Notice
http://mail.csir.co.za/CSIR_eMail_Legal_Notice.html 
 
CSIR Copyright, Terms and Conditions
http://mail.csir.co.za/CSIR_Copyright.html 
 
For electronic copies of the CSIR Copyright, Terms and Conditions and the CSIR
Legal Notice send a blank message with REQUEST LEGAL in the subject line to
Cal...@cs....
This message has been scanned for viruses and dangerous content by MailScanner, 
and is believed to be clean.