Sleepyhead Data Help

had the basic idea... as in the name... but trying to understand them.

see if you can read this - it at least tells how the events are being scored but the copy is bad cause of the scanner (used it on my husband's machine - I run windows 7 - he has XP now and the scanner was for '98 and I have driver for XP but not for anything newer) - I can try to rescan again later if needed

thx gives me something to think about

jules wrote:how is the "snore" value in the summary table determined?

The average is a time Weighted average,
the 90% is the standard statistical 90 (sort the array of data points, take the 90th percentile.)
Both calculations include 0 values, which may or may not be correct, depending on perspective.
Min and max are standard.

On the PRS1 the vibratory snore flags shown in sleep therapy flags in encore, represents all non zero values of this data set. I show it graphed instead. Encore shows a different vibratory snore data set over the waveform display. I show that value in both event flags and on the waveform. Only that one affects pressure.

my data for last night included 7 snores that were reported - 5 of the group were in a period of time where the first and the 5th were 4 min 30 second apart - the other two were 9 1/2 min apart more than an hour earlier

I understand about time weighted averages (taught things like that) - I understand basic statistics (have taught it even) - what data goes into these calculations as for the life of me I can not figure out how 7 events over the time frame that variable snores were reported ended up with a max snore rate of 29

obviously there is data being used that isn't just 1 for snore and 0 for non-snore

I am assuming the units are number/hour - perhaps that is wrong - what are the units?

jules wrote:my data for last night included 7 snores that were reported - 5 of the group were in a period of time where the first and the 5th were 4 min 30 second apart - the other two were 9 1/2 min apart more than an hour earlier

I understand about time weighted averages (taught things like that) - I understand basic statistics (have taught it even) - what data goes into these calculations as for the life of me I can not figure out how 7 events over the time frame that variable snores were reported ended up with a max snore rate of 29

obviously there is data being used that isn't just 1 for snore and 0 for non-snore

I am assuming the units are number/hour - perhaps that is wrong - what are the units?

The Snore data recorded by the PRS1 is confusing. I'm probably restating the same thing as before here.. I'm too foggy to know what's going on today.. But hopefully this clarifies things a little bit.

In SleepyHead, the events flags, flow rate waveform overlay, and Events tab snore events all show the 0x0D (pressure affecting) vibratory snore events.
In Encore, this is shown on the flow rate overlay, plus sleep therapy flags, but only on older/cheaper machines. (Unless it was a bug in the version I saw)

The Snore Graph uses snore "volume" recorded in the graph data (0x11) field (this code also records "total" leaks). This graph is the data source that is currently shown in the details min/max/avg/90 list. Encore shows this in Sleep Therapy Flags on newer machines where this is available. It does not affect pressure.

The peak value of 29 is only happening briefly, and will barely be affecting the weighted calculation. The 0's being included probably aren't helping.
The duration of the snore event is taken from the previous (snore) event to the current one. This is the time multiplier used in (all) the weighted calculations.

As for the units of the graph, I'm only assuming it's related to vibration/volume, and that larger >0 in the snore graph and actually more than smaller >0 values.. Not sure if it's linear, logarithmic or the values recorded here are total B.S.

I occasionally see some PB flagged on my reports but when I look at the actual breaths in the wave flow, while there might be a little waxing and waning it is almost always a far cry from what actual CSR breaths look like.
See this link, bottom of first page. Ozij has inserted examples of the true rounded CSR with some explanation.
viewtopic.php?f=1&t=68516&p=636947#p636947