#1059 "every" keyword incompatible with column header records

[Anonymous]

Either I'm doing it wrong or there's a bug when trying to use the "every" condition in a stacked histogram. If given a dataset with column headers, I have no trouble reading all the even rows. However, if I try to advance one line, to pick up the odd rows, I loose the column headers and the dataset is off. Can someone either show me the right way, or confirm this is a bug?

[Nobody/Anonymous]

test data, script and resulting graph

[Ethan Merritt]

The problem is that the program does the line filtering, both skipping the leading N lines and then every subsequent Mth line, before trying to interpret what remains. That interacts badly with column headers because that header line will be stripped from the input stream before it can be interpreted.

This bad interaction of "every" and "columnheader" is true regardless of the plot style. I'm not sure if the use of histograms introduces any additional problem.

So yeah, that seems to be a bug. I'd have to think about how else the program might handle that combination, or if it's even possible.

[Oblio]

Ok, good, it's not just me. What about hitting the data twice, once with "every ::0::0" that could satisfy the columnheader, then a second call to run over the data? However, if that can work, I don't know the proper syntax.

[Ethan Merritt]

Changing the title to be more informative

[Ethan Merritt]

fix applied for both 4.6 and 4.7