graphs with line and marker without “bridge data”

[arunParticipant]

Hi,

The latest version seems to have made some changes that has affected the results from my old workflow and am curious as to what would be the new way to get back my old results.

I plot data with a large number of points using the simple x-y plot command. In order to distinguish several plot, I use markers. Due to a large number of data points, the marker density is too high. Using a mask and bridging the data results in the plot being drawn to be an approximate and not the original curve (I used this for scientific publications, so I would not like changes to my original data).

My old solution: Plot and x-y graph with only a line (no marker).

Then plot the same data with a marker (no line) and use a mask to control the density of markers.

Provide the same name in the legend for both these graphs. I then get a curve with the line and marker as I like it and the legend shows the combined line and marker identifier.

After the update, I notice that this method results in two plots in the legend.

How may I fix this? I have seen a similar post on “plotting Nth point” here. While that seemed like a good method, the use of the mask results in an approximate line being drawn and not the original.

Thank you

 

[dgteamModerator]

This was a change from a user request to not collapse the commands in the legend into one entry when a mask is used.  Now the entries only collapse in the legend if they are exactly the same data (including checking the mask).

To get the same effect with version 4.6, here is what you can do.

1 – Remove the legend label from command the is drawing your points that has the mask (here it is shown as a Points command, but would be the same for the Plot command).  This leaves just the line in the legend.

 

2 – Next click the gear menu on the Plot command that is drawing all the data and select use in Points command.

3 – Then hide the Points command from view – which removes it from the graph, but not the legend.

Does this help?

 

[arunParticipant]

Thank you for the quick response!

That solution does indeed work well for me, albeit adding an additional step to the workflow.

In fact, the workflow that I described above was a workaround after a new version (in around 2015 may be) changed the original use as mask command. It used to apply only to the markers and the line would remain. That update changed this functionality and I had to discover this new function.

Since you are open to user requests, I request that a feature to easily control the marker density be added.

Datagraph has made my workflow easier and intuitive over the last few years, it would be sad to see it get too complicated.

Thank you again

[arunParticipant]

Hi, I just noticed that this solution does not seem to work if there is a plot offset involved.

Do you have any work around for that?

[arunParticipant]

This method does seem quite convoluted. Are there any plans to provide a more straightforward way to control the maker density as in the previous versions?

Thanks

[dgteamModerator]

We added an option in the beta version to change the stride in the points that should make this much easier. You can download the beta now to use this option.

In the beta, expand the Plot command and there is a menu that says  ‘Every Point’ by default.  Change the menu to ‘Stride’ and an input box allows you to the change the distance between plotted points by entering a value or using a variable as shown in this example.

Does this solution work for you?

(Sorry forgot to post a message that this option was added!)

 

[sandriftParticipant]

I can’t speak for the OP, but this is very helpful and a significant time/complexity savings, thank you.

Could this also be implemented for the Plots command (the OP’s other question)?

Thanks!

[dgteamModerator]

Glad this is helpful!  We recalled you had a similar request @sandrift.

To clarify, we have implemented for the “Plot” command.  Are you asking for the same functionality for the “Plots” command?

 

[sandriftParticipant]

Yep!  I tend to use Plots when I want to show a bunch of spectral lines offset, and although that may seem not to need this (because the combo of  offset and colors might seem to be sufficient) , I try to be considerate of color-blind viewers.  Sometimes different line types (the other obvious alternative) make details of the data I plot harder to see, so having solid lines with symbols would be a great alternative.

[arunParticipant]

This works! “Stride” is exactly what I was looking for!

Thank you very much

[Author]

Posts