SleepHead Statistics anomaly

On the "Statistics" tab in SleepyHead, the "CPAP Statistics" summary lists my AHI for the Last 30 Days as 2.41.

However, the most recent (top) two rows in the "Changes to Prescription Settings" for AHI read: 20 days - 1.89, and 10 days - 1.51.

That doesn't seem to compute, to me. Any explanation for the apparent discrepancy?

AHI calculations in sleepyhead are foobar. See my earlier post.

cnaumann wrote:AHI calculations in sleepyhead are foobar. See my earlier post.

or you're just wrong... (which is more likely)

When I don't understand something, i figure it could just be me; but usually it's not.

or you're just wrong... (which is more likely)

Please check my math.
viewtopic.php?f=1&t=102263&p=957349#p957349

cnaumann wrote:

or you're just wrong... (which is more likely)

Please check my math.
viewtopic.php?f=1&t=102263&p=957349#p957349

how can one check your math when you don't provide the data?

"oh, look, it says something different today?!??!?! and when does the day roll over for you?

. how can one check your math when you don't provide the data?

If you follow the link, you will get to my previous post that includes screen shots.

The screen shots show the total session time. The screen shots also show the total number of events. The screen shots all show the calculated AHI, which does not match the number of events divided by the time.

I can't show you a screen shot from the previous day, because I did not take one. I did record the number on my spreadsheet, but since the number is correct, it is not really interesting.

The total session time in the summary is also incorrect if there are multiple summaries.

Oximeter events are totally foobar as well, it is supposed to flag a change of more than 4% over 8 seconds, but in reality it seems to flag an event every time pulse ox passes 92%, and I suspect other values as well.

And yes, I have been meaning to see if these are already on the bug lists, Submit them if they are not, and offer other support because it really is an amazing, neat and useful program, but it also has quite a few bugs in it still.

cnaumann wrote:

. how can one check your math when you don't provide the data?

If you follow the link, you will get to my previous post that includes screen shots.

I read your gibberage...

tl;dr: it looks fine, but then I come back the next day and stuff has changed!!! and a screenshot, but no data.

upload your sd card to dropbox, and or redo your calculations.

I read your gibberage...

tl;dr: it looks fine, but then I come back the next day and stuff has changed!!! and a screenshot, but no data.

I really have a hard time understanding your level of hostility. If you can't follow the screen shots and the explanation, fine. Don't take it out on me.

cnaumann wrote:

I read your gibberage...

tl;dr: it looks fine, but then I come back the next day and stuff has changed!!! and a screenshot, but no data.

I really have a hard time understanding your level of hostility. If you can't follow the screen shots and the explanation, fine. Don't take it out on me.

stupid gives me hives... sorry...

you're continually posting nonsense, in and among the things you do get right.

you posted detail, said "oh, yeah, it's about right" then came back the next day and said "it changed! bwwaaahhh!!" without posting any detail.

so, simple... back up your allegations, or stfu.

Those screen shots show in order:

The first screenshots shows total PRS1 session time of 5h 17m and a few seconds. That is 5.28 hours

The second screenshot show 52 events, 5 CA, 7 OA and 40 H

That should be an AHI of 9.84 (52/5.28)

The final screen show the reported AHI to be 5.96

And that is clearly wrong.

Why are you so fixated on my claim that the AHI numbers changed from what was reported when I first imported that data?


I though for a while that you were simply terse and pedantic, but it is much more than that. You are a classic bully. You know quite a bit, and you have been helpful to me and others and I am grateful for that. But you have some serious issues and you are probably driving away more people than you are helping.

yes, I bully stupidity, and stamp out ignorance... but I think you've got more than one person could handle.

since you're unable to figure out how to diagnose the problem on your own... and you haven't uploaded your sd card so someone else can figure it out for you, lets try this...

what's the difference in the session list between your screenshot where it showed 5.28 hours, and the last screenshot where it showed eight something hours?

cnaumann wrote:

I read your gibberage...

tl;dr: it looks fine, but then I come back the next day and stuff has changed!!! and a screenshot, but no data.

I really have a hard time understanding your level of hostility.

Yup... so do many of us! If you read PR's posts, you'll see that lots of people feel the same way.

Well if you break out your calculator and add up the session times, I think you will find that the 8h20m minute time is wrong. The session times do not add up to that. The session times shown on the first screen shot are all there are, and I believe they are correct.

There is an oximeter session that lasts 8h10m but it should not be used in the calculation of AHI. It may be part of the problem, but It is also not 8h20m.

It gets stranger. When I go to the overview statistics, the bar graph for usage on the 20th goes almost to the 12 hour mark. However, when I highlight it, it show 8.73h. Neither of those times is correct, nor are they consistent with the 8h20m show on the daily screen.

In fact, when I highlight any bar graph in the usage window it gives an incorrect summary if there are multiple sessions.

No one has asked me to upload an SD card. I don't even know how to do that. Several people have asked about how AHI is calculated over multiple sessions. Several people responded with how it should[\i] be calculated. I responded that there are some issues with the calculation and showed some screenshots. What have I done that is so ignorant or unreasonable?

I just realized that 8h20m could be the total of the oximetry session and all non-overlapping PRS1 sessions. I bet the AHI number changed after I imported the oximetry data and I did not notice it until the next day...

[quote="palerider"]yes, I bully stupidity, and stamp out ignorance... [\quote]
And replace it with your own.