Support for Extrapolated Values on Timeseries - LOCF Interpolation

Hi Akeel,

On what type of time series data is this behaviour required?  Is this about carrying forward the value of a state representation (i.e. off / on, or open / closed) or does it also related to measurements / numeric type data?

Hei ​@akeelishak just following up on this request . Would you want this to be applied to calculations too ? is this for cases where the sensor is maybe down for a while / is delayed in its response to the extractor ?

Hi , ​@Glen Sykes ,  ​@Arun Arunachalam 

I also believe this feature is necessary.
In our case, we register manually conducted analysis results in CDF,
and this data is updated only about once every two weeks.
With the existing “Interpolation” feature, only interpolation between data points is possible, so the latest data is always missing.
This is inconvenient when performing calculations on Charts using this data.
What we want to achieve is illustrated in the figure below.

While not a proper solution, could it be a workaround to add a “dummy” datapoint in the far future, say in year 2099? Especially since this is a step time series.

Hi , ​@matiasholte 

Thank you for your response.
By adding a dummy data point, interpolation would work, so I believe it would achieve what we want.
However, registering a dummy data point within the correct data—even if it is clearly identifiable—could potentially lead to unexpected issues, so it might be a bit challenging for us.
That said, the idea of such an approach was something we hadn’t considered, so we truly appreciate your valuable suggestion.