search

automeris-io / WebPlotDigitizer

Computer vision assisted tool to extract numerical data from plot images.
https://automeris.io
GNU Affero General Public License v3.0
2.58k stars 354 forks source link

x stpe #284

Closed Babizinho closed 2 years ago

Babizinho commented 2 years ago

On my simple xy graph, "averaging window" works fine, but I need y readings for integer x values and "x step" doesn't work. When I click "Run", nothing happens. I have tried changing x step value and other changes, to no avail. How do I get "x step" to actually run?

Babizinho commented 2 years ago

mistyped title, will re-enter post...

nbehrnd commented 2 years ago

May you illustrate the issue by submission of an additional .png with about the diagram in question?

Babizinho commented 2 years ago

Thanks for prompt response. Here's the graph image... AGe of onset CENTRE

nbehrnd commented 2 years ago

Contrasting to the abscissa (the horizontal axis), the ordinate (vertical dimension) lacks tics, the little lines of graduation. Given the style of the font on the axes which are lining (contrasting to old style figures like the year on a penny with different base and vertical dimension (Wikipedia)) my assumption these (missing) tics were approx. at the level where where the downward stroke of the «5» and the arc meet. (Perhaps screen photo 02 (zip archive below) illustrates this better.) Because this is an approximation, my recommendation is: if you happen to have access to the original plot with tics on both axes, repeat the digitization.

The default averaging digitization works, as described by you. For the one with x = 0, 1, 2, 3, ... you are going to need to adjust the levels of y_min and y_max to extend beyond the interval y = [0...1]. Once set, e.g. y_min = - 0.2 and y_max = 1.2, the digitizer works just as well as one is used to.

dataset.zip

nbehrnd commented 2 years ago

Note: For some reason, you already closed this thread some hours ago, yet later added the plot in question. On the other hand this issue, likely equally aiming to resolve the same question. Consider to close the other one.