Graphing Vegetation Biomass Data with Kepler: A Step-by-Step Exercise

1. Kepler Exercise Deana Pennington University of New Mexico January 9, 2007

2. Graphing data Vegetation biomass test data • With EML • Without EML

3. Sevilleta LTER Biomass Plots Biomass = f(species, cover%, height, season)

4. This Exercise • Using a subset of the data for 1 species in 1 quadrat, extract vegetation cover% and height, and create graphs of those values through time

5. Get data file • Copy from tundra://training_lab/downloads veg_data_test.csv vegdataDP.xml vegdata2DP.xml • Copy to c:\kepler\workflows\eco

6. View the end result • Open vegdataDP in kepler • File->Open->vegdataDP.xml • Run

7. With metadata (easier) • Open a new graph editor in Kepler File->New->Graph Editor • Switch to the Data tab • Remove KU Digir and Geon from Sources • Search for “test” • Add “Vegetation Test Data” to canvas Drag and drop

8. Explore the dataset • Look at the metadata Right-click on icon-> Get Metadata • Tool tips on ports

9. Director • Add a director Click on the Components tab Double-click on Components Double-click on Directors Drag and drop the SDF Director onto the canvas • Configure director Double-click or right-click->configure Change iterations to 9 (the number of rows to be read in the file) Click Commit to save

10. Display the file • Search for “display” • Find the Display actor (last icon) • Drag and drop the Display actor onto the canvas

11. Hooking up ports • Find the “cover” port Use tool tips Note double type • Click-hold the cover port and drag to the input port on the Display actor • Run the workflow Click the teal triangle or • Delete the Display actor

12. Add graph • Search for “plotter” • Add XY Plotter actor to canvas • Hook up cover and height output ports to graph input ports • Ru

13. Run the workflow • Note lousy plot format • Close the display • Change the plot format Click on plot format icon

14. Reformat the plot • Change Title to something meaningful (I used Vegetation Volume) • Add axis labels • Select dots • Unselect Connect • Apply • Close Run Window • Run from run button • Save your work File->Save-> c:\kepler\workflows\eco\veg_data_xxx.xml, where xxx is your initials

15. Without EML • The harder way, without EML-described metadata

16. Cheat-look ahead Open vegdata2DP.xml from Kepler->File->Open File-> [locate file] Opens the file Separates file into columns Extract cover and height columns Converts to decimal number Converts array to elements Constructs graph Note: all this would not be necessary if EML used

17. Start a new workflow • Open the workflow editor File->New->Graph Editor • Select a director Components Directors Drag and drop SDF Director • Configure director Right click on director Change number of iterations from 0 to 1 • Commit

18. Read the file • Browse Components=> Data Input => Local input; drag and drop Read Table • From General Purpose, get two copies of the Expression actor • Configure Expression 1 to hold the path and file name for the veg_data_test.csv file • Configure Expression 2 to specify the delimiter “,” • Link the output port of Exression 1 to the file name port of Read Table • Link the output port of Expression 2 to the delimiter port of Read Table • Add the Display actor and hook it up to the string output port • Open Read Table and remove the “pairs” statement at the end • Change the R working directory to your directory • Uncheck the graphical display check box • Run the workflow – it should show you the file contents

19. Plot the columns Get the R Expression actor from General Purpose Right-click Configure Ports Add an input port named df1

20. Plot the file (cont’d) • Hook up the dataframe output port from Read Table to the new input port • Configure the actor – change the R script to specify the cover and height columns (as shown below) and to plot

21. Plot the file (cont’d) • Get Image J • Hook the graphics file output port of R Expression to the input port for Image J • Run workflow