You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The getStatistics() method is buggy. We sometimes need a way to auto-scale plots upon load. Scaling is easy if it's user-driven, as in Fluxtream when you click on the double-headed arrow to maximize the plot height. But it appears to be much harder to do well upon loading a channel. The getStatistics() method has an optional callback argument, but the callback usually gets called too soon, before all the tiles have actually loaded, so the stats are useless. And we don't want to just use the min and max values in the channel spec (e.g. as returned by https://fluxtream.org/api/bodytrack/users/USER_ID/sources/list) because those min/max are for the entire time range of the channel's data, and not just for the small window of time Paul's loading up. Same is true for the "default_graph_specs" API method (e.g. https://fluxtream.org/api/bodytrack/users/USER_ID/sources/DEVICE_NAME/default_graph_specs).
The text was updated successfully, but these errors were encountered:
It's been a while since I've looked at that code, but it looks like the bug is in the definition of meetsFilter in AllLoadEventListener. This method immediately assumes that anything it's asked about has loaded, by removing it from the list of things left to load. Probably the fix is to change FilteredEventListener to have an onLoad(TileDescription) method, so meetsFilter only has one job.
The getStatistics() method is buggy. We sometimes need a way to auto-scale plots upon load. Scaling is easy if it's user-driven, as in Fluxtream when you click on the double-headed arrow to maximize the plot height. But it appears to be much harder to do well upon loading a channel. The getStatistics() method has an optional callback argument, but the callback usually gets called too soon, before all the tiles have actually loaded, so the stats are useless. And we don't want to just use the min and max values in the channel spec (e.g. as returned by https://fluxtream.org/api/bodytrack/users/USER_ID/sources/list) because those min/max are for the entire time range of the channel's data, and not just for the small window of time Paul's loading up. Same is true for the "default_graph_specs" API method (e.g. https://fluxtream.org/api/bodytrack/users/USER_ID/sources/DEVICE_NAME/default_graph_specs).
The text was updated successfully, but these errors were encountered: