Open exscalliber opened 11 months ago
More context here from me:
From my understanding, Reveal only implements CategoryXAxis
or possibly OrdinalTimeXAxis
ATM.
It should also give an OPTION to use TimeXAxis
, and possibly CategoryDateTimeXAxis
for the time series specifically. TimeXAxis
has some hoops to jump through since it supports different labels at different visible time ranges, so may need some UI work from Reveal. The observed behavior is actually expected and desired behavior for CategoryXAxis
(we WANT it to remove areas where there is no data). But to support the desired behavior here they need a true continuous time scale on the x axis like TimeXAxis
or CategoryDateTimeXAxis
.
CategoryDateTimeXAxis
is the quicker hit, since it doesn't automatically support different date/time formats at different visible time ranges, so shouldn't require altering the UI in Reveal that allows for selection of the date format.
SDK Version
1.6.1
Client Framework
HTML/JavaScript
Server Platform
None
Operating System
Linux
Description
When you have data with irregular timestamps, the ticks between the dates are completely incorrect. In the attached images, i have the visualization from reveal that shows the incorrect spacing, and the same exported visualization from excel that shows the time series data with correct spacing/ticks between times. I've even found in the infragistics graphs that the line chart shows the same error, although this looks like a case of bad data which I've also attached (https://www.infragistics.com/products/ignite-ui-react/react/components/charts/types/line-chart#line-chart-data-structure). The reveal visualization should do exactly what excel does and fills in the dates between the ticks otherwise the data gets misrepresented when there are long time gaps between data. This issue affects these charts: Time Series, Stacked Area, Area, Line, Step Area, Step Line, Spline area, Spline, Combo.
Reveal Visualization
The reveal visualization exported to Excel showing correct spacing
Infragistics spacing issue (somewhat related)