Contents	Previous	Next

CHAPTER 5 Plotting CDMS data in Python

5.1 Overview

Data read via the CDMS Python interface can be plotted using the vcs module. This module, part of the Climate Data Analysis Tool (CDAT) is documented in the CDAT reference manual. The vcs module provides access to the functionality of the VCS visualization program.

Examples of plotting data accessed from CDMS are given below, as well as documentation for the plot routine keywords.

5.2 Examples

In the following examples, it is assumed that variable psl is dimensioned (time, latitude, longitude). psl is contained in the dataset named 'sample.xml'.

5.2.1 Example: plotting a gridded variable

1 import cdms, vcs
2 #
3 f = cdms.open('sample.xml')
4 psl = f.variables['psl']
5 sample = psl[0]
6 w=vcs.init()
7 #
8 w.plot(sample)
9 f.close()

Notes:

Line Notes

5 Get a horizontal slice, for the first timepoint.

6 Create a VCS Canvas w.

8 Plot the data. Because sample is a transient variable, it encapsulates all the time, latitude, longitude, and attribute information.

9 Close the file. This must be done after the reference to the persistent variable 'psl'.

Thats it! The axis coordinates, variable name, description, units, etc. are obtained from variable sample.

What if the units are not explicitly defined for psl, or a different description is desired? plot has a number of other keywords which fill in the extra plot information.

5.2.2 Example: using plot keywords.

w.plot(array, units='mm/day', file_comment='High-frequency
reanalysis', long_name="Sea level pressure", comment1="Sample
plot", hms="18:00:00", ymd="1978/01/01")

Note: Keyword arguments can be listed in any order.

5.2.3 Example: plotting a time-latitude slice

Assuming that variable psl has domain (time,latitude,longitude), this example selects and plots a time-latitude slice:

1 samp = psl[:,:,0]
2 w = vcs.init()
3 w.plot(samp, name='sea level pressure')

Notes:

Line Notes

1 samp is a slice of psl, at index 0 of the last dimension. Since samp was obtained from the slice operator, it is a transient variable, which includes the latitude and time information.

3 The name keyword defines the identifier, by default the name in the file.

5.2.4 Example: plotting subsetted data

Calling the variable psl as a function reads a subset of the variable. The result variable samp can be plotted directly:

...
1 samp = psl(time=(0.0,100.0), longitude=180.0)
2 w = vcs.init()
3 w.plot(samp)

5.3 plot method

The plot method is documented in the CDAT Reference Manual. This

section augments the documentation with a description of the optional key

word arguments.

The general form of the plot command is:ch4_cdms_4.0.html/#4.1_Overview

canvas.plot(array [, args] [,key=value [, key=value [, ...] ] ])

where:

• canvas is a VCS Canvas object, created with the vcs.init method.
• array is a variable, masked array, or Numeric array having between two and five dimensions. The last dimensions of the array is termed the 'x' dimension, the next-to-last the 'y' dimension, then 'z', 't', and 'w'. For example, if array is three-dimensional, the axes are (z,y,x), and if array is four-dimensional, the axes are (t,z,y,x). (Note that the t dimension need have no connection with time; any spatial axis can be mapped to any plot dimension. For a graphics method which is two-dimensional, such as boxfill, the y-axis is plotted on the horizontal, and the x-axis on the vertical.

• args are optional positional arguments:

• key=value, ... are optional keyword/value pairs, listed in any order. These are defined in Table 5.1 on page 145.

  key                                type                            value

	Contents	Previous	Next