I have a huge data table of 2705 rows * 2106 columns, that looks basically like this:
RowID A_x A_y B_x B_y C_x C_y
0 1 10 1 11 1 13
1 2 19 2 21 2 18
2 3 40 3 45 3 37
3 4 75 4 83 4 79
4 5 102 5 109 5 100
I would like to create individual x,y plots for each object (A, B, C etc.), basically reproducing time-data curves produced by an instrument for each object. As can be seen the x values are the same for all objects such this might be used to advanatge.
Any suggestions for a reasonable fast workflow using JFreeChart are highly appreciated!
if you can work with one x column (say A_x) for all of your plots you just need to exchange the way of looping.
the column loop start will output one of the included columns in each run.
even more simple, without need to transpose the whole dataset (which can be problematic when table are large). As above, identify th n objects that are documented, loop over to create the n itterations needed (here, n=3). The trick is to use the filter ref column node, with columns that are created at each itteration, with the right header: obtained by transposing the three variable created based on "object" name. each loop, the two columns containing the data documenting the object are fed to jfree line chart...concatenated in a table.
one possible option is to extract the names of the parameters which are described by the x and y values.
extract col header, transpose, split, group by give A,B and C. One loop over A to C allow to create param_X and Param_Y columns. Intermediate transposition helps to filter parameter_X and parameter_Y with a filter ref row node. once the two cols are available, they are concatenated and transposed back to create the two cols param_X vs param_Y which are fed to the J free line chart. Loop dependant variable can be built to illustrate axis, and to build the graph title.
each generated graph is annotated and fed to a table. At will,, you can also switch to save graph node and use title graph as name (you'll be needing to store the path to the file as additional text in the garph title though)
enclosed a snapshot of the WF and the workflow itsefl, as well as the final table.Note that I did use the data you provided and that I did add some extra values for A and B.
hope this will do the job.
Many thanks Bruno, this was extremely helpful!