Pre loader

1

Welcome to the SciChart Forums!

  • Please read our Question Asking Guidelines for how to format a good question
  • Some reputation is required to post answers. Get up-voted to avoid the spam filter!
  • We welcome community answers and upvotes. Every Q&A improves SciChart for everyone

WPF Forums | JavaScript Forums | Android Forums | iOS Forums

All Answers » Re: Heatmap missing data points » Comments for "Re: Heatmap missing data points"
  • MIha Rozina
    It shouldn’t be the same as my previous question. This time the size of the data series is much smaller than the number of pixels rendered. So the number of texture pixels should be equal or smaller to the number of pixels visible on screen. Note that SciChartSurface is 200 px in width, while the data series is only 116 points in width. Please reconsider your answer and revisit the question. Thank you.
  • Andrew Burnett-Thompson
    Apologies, I see the difference now. OK our dev team is off today for a public holiday, but I will make them aware to investigate this on Monday. Kind regards, Andrew
  • MIha Rozina
    Great, thank you!
  • MIha Rozina
    I hope this question was not forgotten :)
  • Yura Khariton
    I think this issue is caused by same limitation of heatmap implementation which I described in question mentioned by Andrew. For now heatmap is rendered as single texture which has size of the data series ( in your case it has size 116×50 pixels ). Width of chart is 200 pixels minus width of yAxis which I would say takes ~20% of chart’s width on your screenshot. So let’s say that width of heatmap is ~160 pixels. This means that for each texture pixel we have 1.38 device pixels and on real device you can’t render such amount of pixel for obvious reasons ( you can set red color for 1 or 2 pixels but you can’t set it for half of the pixel ). So in this case depending on coordinate of texture pixels OpenGL rasterizer in one case sets 1 pixel (e.g. case when texture pixel should be drawn from 1 to 2.38 pixel it will be rounded and drawn from 1 to 2 pixel ) and in other sets 2 pixels to be red ( e.g. case when texture pixel should be drawn from 1.2 to 2.58 device pixels which after rounding to closest value will become 1 to 3 ). That’s why you see different line thickness. Unfortunately there is nothing we can do to improve this situation without rewriting heatmap to draw each device pixel on our own instead of leaving interpolation of heatmap texture pixel values to OpenGL.
  • MIha Rozina
    I understand that concept, but the problem with my example is that the red line is not drawn at all! Not that one red line is 2 px width and one is 1 px width. That is both fine and I accept the limitation. The problem is that the first red line is 0 px in width. Based on the DataSeries there should be a red line at x = 0, which is not seen on the screenshot. That is my complaint and I don’t feel it has been addressed thus far by your comment.
  • Andrew Burnett-Thompson
    I have a suggestion for Yura K which I will discuss with him: about using Linear Interpolation. Will get back to you