Timeline for Comparing sensor data across multiple experiments
Current License: CC BY-SA 4.0
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Aug 18, 2019 at 19:44 | answer | added | EdM | timeline score: 1 | |
| Aug 8, 2019 at 19:36 | history | edited | EdM | CC BY-SA 4.0 |
Added critical information from a comment and reorganized a bit.
|
| Aug 8, 2019 at 19:24 | comment | added | Floris | Could I, for example, take the average over every minute, and test that? That would already reduce the amount of tests by a lot (18000 to 30). During the experiment (total duration 30 mins), a force is applied for the first 10 or so minutes, after which it remains constant. That's why simply taking the mean of all 18000 isn't viable - the conditions aren't the same across all 18000 timepoints | |
| Aug 8, 2019 at 17:01 | comment | added | EdM | Doing 18000 separate tests would lead to a big problem with multiple comparisons. You need some way to combine the 18000 measurements in each series--if the mean alone can't be used for some reason, then some type of time-series modeling. Accounting for changes over time might also resolve your normality issue, as might some data transformation like log or square root. More information on the nature of these data series might get you a more useful answer. | |
| Aug 8, 2019 at 16:00 | review | First posts | |||
| Aug 8, 2019 at 16:06 | |||||
| Aug 8, 2019 at 15:58 | history | asked | Floris | CC BY-SA 4.0 |