Overview
You have seen already how to visualise your outputs either on the display or save it to file. Most processing you will do with Metview will lead to an visualisation you might want to save to either publish it on web pages or in reports or simply to keep records. This session will give you more information on how you can save and customise your visualisations.
Output formats
All graphical output formats are generated through ECMWF's graphics library Magics. This means that all graphical capabilities in Metview depends on what Magics offers. Detailed descriptions of the output formats and their settings can be found in the Magics reference documentation.
The table below gives you a list of all the formats Magics/Metview support and how you can use them.
Format | parameter | How to visualise | What to use for |
---|---|---|---|
Qt | Metview display window | Interactive usage within Metview | |
PostScript | ps/eps | okular, ghostscript/gv | Printing, publications |
okular, acroread | Web, archiving | ||
PNG | png | web browsers, display, xv | Web, presentations -> animations |
SVG | svg | web browsers, inkscape | Web HTML5, editing with inkscape |
KML/KMZ | kml | Google Earth, Google Maps, OpenLayers | Interactive publications |
Metadata
When a large amount of plots is generated it is often hard to find later plots with specific contents. What can help is to store additional information with the plots to describe what the content is. This descriptions about the content are called Metadata. Magics/Metview support the saving of such metadata when the format allows this. Especially in text/XML based formats, such as Postscript and SVG, you can use simple UNIX tools like grep to search the files for specific keywords.
How to save your visualisations
Here we quickly recap how you can save your display from the interactive plot window or save them within Macros.
Exporting from the plot window
Whenever you have a plot window open which displays any map or graph, you are able to export the plot into other file formats. You can either select the export button in the menu (looks like a disk)
, by selecting File > Export or by using the Ctrl+s keyboard shortcut.
You will get a dialogue as shown below
Tasks
- From an open display window, try to save your plot as a PDF.
- Now try to save your plot as a SVG without and with fixed dimensions.
- Open both files in Firefox and see how the plots behave when you resize the browser window.
- Open one of the SVGs in inkscape and edit the images (for example add some text) and save it as a PDF.
Setting output formats in macro code
You have already seen in previous exercises how to save plots from macros.
The following code example shows how to set various parameters for the different output formats. Note that in most cases the default values are sufficient.
# # Setting common output options for multiple formats # output_common = ( output_name : "/tmp/myOutputTest", # specify full path output_width : 1000, # set width in pixel output_title : "Map of Z500", # give a title to be used by a viewer output_debug : "ON", # print extra debug information into file output_filelist : "ON", # record the list of output files generated output_filelist_name : "/tmp/filelist.txt" # specify where the list of generated files goes ) ps = ps_output( output_common, output_ps_scale : 0.9, # scale content to 90%, this is useful for some printers output_ps_colour_model : "CMYK" # set colour model to CYMK (used by printers) ) png = png_output( output_common, output_cairo_transparent_background : "ON" # generate PNGs with transparent background ) svg = svg_output( output_common, output_svg_fix_size : "ON", # normally SVGs are scalable - this fixes the size to 'output_width' output_svg_meta : "Metview map of Z500", output_svg_desc : "This file was generated for the Training course" ) kml = kml_output( output_common, kml_description : "This file was generated for the Training course", kml_author : "Stephan Siemen", kml_link : "http://www.ecmwf.int", kml_latitude : 30, # latitude to where Google earth centres the view kml_longitude : 120, # longitude to where Google earth centres the view kml_coastlines : "ON" # normally ) output_drivers = [ps, png, svg, kml] setoutput(output_drivers) data = read("z500.grib") plot(data)
Setting output dependent on runmode
The way to code run mode dependent outcomes is by using the function runmode(). It returns a string with the run mode:
mode = runmode()
So it is enough to check this string and to code accordingly, either using if/else conditions or the case/of condition test. To introduce the new functionality replace the existing unconditional call to setoutput() by the following lines of code :
# check run mode mode = runmode() # select outcome dependent on run-mode if(mode = "execute") then setoutput(to_pngfile) else if (mode = "batch") then setoutput(to_psfile) else if (mode = "visualise") then print('Plotting to screen') else if (mode = "prepare") then print('Plotting to screen') else fail("Only execute, batch and visualise allowed") end if
Now, depending on how you call the macro your output will be directed to different media. Choose different options from the icon’s right-click menu to see what happens. Note that you can also simulate these actions from within the Macro editor ( Program | Run Options). The ‘prepare’ run mode is the default one when you run your macro from the Macro editor. The ‘batch’ run mode will be explained in a later paragraph.
If you select an option not covered by the allowed run modes (e.g. Save or Examine), the macro will stop, turn red (failed run) and issue an error message - this behaviour is provided by the fail() function. A related function, stop(), will do the same but allow the macro to exit in the green state (successful run). Note that you may have to see that it has worked! To delete the output files before running the macro in order to
To run the macro in batch mode, you call Metview with the option -b followed by the macro name on the command line (assuming you are running from the same directory as the macro - otherwise you must provide a path to it). For example:
metview -b mymacro.mv
The newlines within the conditional branching part of the code are down to personal preference. You could also have formatted the code as follows:
if (mode = "execute") then
setoutput(to_pngfile)
else if (mode = "batch") then
setoutput(to_psfile)
If You Have Extra Time ...
One important piece of functionality that we have omitted is the ability to save our derived data. Add some code to handle the Save run mode by saving the derived data in a file. There are three things you will have to do:
- at the end of the run mode checks, instead of
else fail ( ... )
- we should not fail if the mode is “save”:
else if (mode <> "save")
then fail ( ... )
- we must again check the run mode, and if it is Save then write the data to a file; otherwise plot the data.
Pages
Some graphical formats, such as PostScript and PDF, allow multiple pages within the documents. Other formats, such as PNG, which will contain a single page at the time and therefore contain a number in their name to indicate which page number they contain. You can trigger a new page in Metview Macro with the function
newpage()
This function is normally used within loops to generate output of each iteration on a separate page (or file).
Task
Take one of your existing Macros containing a loop and try to print each iteration on a page.
Further processing outside Metview
Of course you can do further processing of plots outside Metview. The page Generating animated GIFs from Metview plots gives you some helpful advice if you want to build animations.
Metview/Magics also supports special tags in SVG for the Inkscape graphical editor. This open source editor is great for when you need to further annotate your maps.
KML as output format
KML is a very special output format. It contains the it has no notion of output size. KMZ is the compressed version of the KML files and is written by default.You can write out the uncompressed KML if you want to debug the output.
Be careful
The KML output is till experimental and we look for feedback on it. KML can only be generated if the Cylindrical projections is selected!
Using Web Map Services (WMS)
Metview provides a module to request maps from a Web Map Services (WMS). The WMS standard is defined by the Open Geospatial Consortium (OGC), which is very popular in the GIS community. The WMS module can be used to query an OGC-compliant WMS server, retrieve maps from it and overlay them with other data. There is a separate tutorial on how you can integrate maps in Metview.The only thing you need to know is that to use a map service you need to request a so-called GetCapabilities document first. This document contains all information (in human and machine readable form) on what maps are offered and what the client can do. Metview will parse this file and offers you a convent and easy to use interface to request the maps you want.
Extra tasks
- Work through the WMS tutorial. You can skip some parts, like the manual editing for request if you wish so.
- Try the following WMS services:
- NASA: http://neowms.sci.gsfc.nasa.gov/wms/wms
- NOAA: