2D Plotting with PGPLOT

Welcome to this tour of the PDL's PGPLOT interface.

This tour will introduce the PDL's PGPLOT plotting module and show what this powerful package can provide in terms of plotting. It is not designed to give a full tour of PGPLOT, you are advised to see the routines provided with pgperl for that.

The PDL::Graphics::PGPLOT module provides a high-level interface to PGPLOT. However if you want even better control of your plots you might want to include the PGPLOT module specifically:

    use PGPLOT;

One aspect of PGPLOT that requires mention is the use of devices: Normally PGPLOT will inquire you about what device you want to use, with the prompt:

    Graphics device/type (? to see list, default /NULL):

pdl> # ensure the module is loaded (required for PDL versions>= 2.004)

pdl> use PDL::Graphics::PGPLOT;
pdl> # The size of the window can be specified

pdl> $ENV{PGPLOT_XW_WIDTH}=0.3;
pdl> # You can set your device explicitly

pdl> dev('/XSERVE');

pdl> # First we define some variables to use for the rest of the demo.

pdl> $x=sequence(10);
pdl> $y=2*$x**2;
pdl> # Now a simple plot with points

pdl> points $x, $y;

Output

pdl> # Here is the same with lines
pdl> line $x, $y;

Output

pdl> # If you want to overlay one plot you can use the command
pdl> # 'hold' to put the graphics on hold and 'release' to
pdl> # revert the effect
pdl> points $x, $y, {SYMBOL=>4};  # The last argument sets symboltype
pdl> hold;
pdl> # Now draw lines between the points
pdl> line $x, $y;
pdl> # Plot errorbars over the points
pdl> $yerr=sqrt($y);
pdl> errb $x, $y, $yerr;
pdl> # To revert to old behaviour, use release
pdl> release;

Output

pdl> bin $x, $y;
pdl> # This plots a binned histogram of the data and as you can
pdl> # see it made a new plot.

Output

pdl> # 2D data can also easily be accomodated:
pdl> # First make a simple image
pdl> $gradient=sequence(40,40);
pdl> # Then display it.
pdl> imag $gradient;
pdl> # And overlay a contour plot over it:
pdl> hold;
pdl> cont $gradient;
pdl> release;

pdl> # PDL::Graphics::PGPLOT contains several colour tables,
pdl> # a more extensive collection can be found in 
pdl> # PDL::Graphics::LUT
pdl> #
pdl> # (note: the call to lut_names() can take a few seconds to execute)
pdl> #
pdl> use PDL::Graphics::LUT;
pdl> @names = lut_names();
pdl> print "Available tables: [ ", @names, " ]\n";
pdl> # use the first table
pdl> ctab( lut_data($names[0]) );
pdl> use PGPLOT;
pdl> pglabel "", "", "Colour table: $names[0]";

Available tables: [ aips0 backgr bgyrw blue blulut color green heat
idl11 idl12 idl14 idl15 idl2 idl4 idl5 idl6 isophot light manycol pastel
rainbow rainbow1 rainbow2 rainbow3 rainbow4 ramp random random1 random2
random3 random4 random5 random6 real red smooth smooth1 smooth2 smooth3
staircase stairs8 stairs9 standard ]

Output

pdl> # To change plot specifics you can either use the specific PGPLOT
pdl> # commands - recommended if you need lots of control over your
pdl> # plot.
pdl> #
pdl> # Or you can use the new option specifications:
pdl> # To plot our first graph again with blue color, dashed line
pdl> # and a thickness of 10 we can do:
pdl> line $x, $y, {COLOR=>5, LINESTYLE=>'dashed', LINEWIDTH=>10};

Output

pdl> # Now for a more complicated example.
pdl> # First create some data
pdl> $a=sequence(360)*3.1415/180.;
pdl> $b=sin($a)*transpose(cos($a));
pdl> # Make a piddle with the wanted contours
pdl> $contours=pdl [0.1,0.5,1.0];
pdl> # And an array (reference to an array) with labels
pdl> $labels=['A', 'B', 'C'];
pdl> # Create a contour map of the data - note that we can set the colour of
pdl> # the labels.
pdl> cont($b, {CONTOURS=>$contours, linest=>'DASHED',
pdl> LINEWIDTH=>3, COLOR=>2, LABELCOL=>4});
pdl> hold;
pdl> pgqlw($linewidth);
pdl> points $a->slice('0:-1:4')*180./3.1415;
pdl> release;

Output

pdl> #
pdl> # More examples of changing the plot defaults
pdl> # 
pdl> $a = 1+sequence(10);
pdl> $b = $a*2;
pdl> $bord_opt = { TYPE => 'RELATIVE', VALUE => 0.1 };
pdl> line log10($a), $b, { AXIS => 'LOGX', BORDER => $bord_opt };

Output

pdl> #
pdl> # We can also create vector maps of data
pdl> # This requires one array for the horizontal component and
pdl> # one for the vertical component
pdl> #
pdl> $horizontal=sequence(10,10);
pdl> $vertical=transpose($horizontal)+random(10,10)*$horizontal/10.;
pdl> $arrow={ARROW=> {FS=>1, ANGLE=>25, VENT=>0.7, SIZE=>3}};
pdl> vect $horizontal, $vertical, {ARROW=>$arrow, COLOR=>RED};
pdl> hold;
pdl> cont $vertical-$horizontal, {COLOR=>YELLOW};
pdl> release;

Output

pdl> #
pdl> # To draw [filled] polygons, the command poly is handy:
pdl> #
pdl> $x=sequence(10)/5;
pdl> poly $x, $x**2, {FILL=>HATCHED, COLOR=>BLUE};

Output

pdl> #
pdl> # the latest feature of PDL are complex numbers
pdl> # so let's play with a simple example
pdl> #
pdl> use PDL::Complex;
pdl> $z50 = zeroes(50);
pdl> $c = $z50->xlinvals(0,7)+i*$z50->xlinvals(2,4);
pdl> line im sin $c; hold;      # the imaginary part
pdl> line re sin $c;            # real
pdl> line abs sin $c; release;  # and the modulus

Output

pdl> #
pdl> # more complex numbers
pdl> #
pdl> use PDL::Complex;
pdl> $c =  zeroes(300)->xlinvals(0,12)+i*zeroes(300)->xlinvals(2,10);
pdl> $sin = sin $c;
pdl> line $sin->im, $sin->re;   # look at the result in the complex plane

Output