arun

Hello, Let's say I have a multi-series line chart with plots that overlap for some range of data. If I want to click on a data point for a specific plot that is covered by another plot, there is no way to do so. Please add some mechanism to allow me to select a specific point when plots overlap. Thanks, Arun

Hello, Let's say I have a multi-series line chart with plots that overlap for some range of data. If I want to click on a data point for a specific plot that is covered by another plot, how can I do that? Is there some workaround for this scenario? Thanks, Arun

It's difficult (but not impossible) to click 1-values on a log column chart. Perhaps you could make it possible to click on the count label ("1")? It's also difficult to discern the short column. I've had to alter the Y-axis range to dip down to 0.5 or so, even though a count of less than 1 is nonsensical. Thanks, Arun

There are at least a couple issues with log plots: 1) If the maximum y-value is 1, the plots for the 1-values are up at the top of the chart, well above the title. 2) With the LogColumn2D chart, if there are no non-zero values, the chart causes the browser (at least Firefox) to hang. Fortunately Firefox offers to kill the chart movie after a minute or so. I would expect at least an empty chart to be rendered. Bonus enhancement request: 3) It's difficult (but not impossible) to click 1-values on a log column chart. Perhaps you could make it possible to click on the count label ("1")?

When we have multi-series bar charts, we tend to use stacked columns in order to save space and convey a total for the column. Understandably, stacked bar charts aren't possible with logarithmic scales, so we're presently stuck with multi-series side-by-side charts. What would be nice would be a way, for charts using a logarithmic scale, to overlay each of the multi-series columns at a x-axis position so that space is conserved. I think ideally the overlay would be shifted slightly for each series so that each of the columns would be visible even if there were two columns with the same height. In addition, the columns should be sorted in order of y-value, with the largest column on the bottom and the smallest on the top. Perhaps a 3D bar chart with a log scale would achieve similar results. Thanks, Arun