SleepyHead 0.8.8 "Almost Beta" Release

yardbird wrote:Linux
Kubuntu 11.10 64-bit

Also need to: sudo apt get install build-essential in order to have the g++ libs and stuff necessary to build from source

IntelliPap Auto with Smartflex

Was able to read the SD card after figuring out that the dropdown in the dialog box doesn't show me my USB-connected SD card reader. I had to go down to where you select directory. Once read, it pulled in information from the card.

I've got 960 days on this machine so the Overview is pretty tight with data. Printing it would be less than useful as the graphs are too tight to be readable. It will be better when you can print the daily stats/graphs.

I could not find Smartflex data if it's in there anywhere.

Thanks for testing the Intellipap stuff.. It's still a little young.. and still buggy.

You can zoom in/select an area and when you print, it will only print the graphs as you see on the screen.

Also, daily printing is done..
Hint.. File->Print.

Edit: as for smartflex data, it's recorded.. I still need to add it to the settings overview graph..
Still haven't shown the setting data for the other two machines.. Having the extra machine is extra incentive to do it properly

Pugsy wrote:Up and running with 0.8.8 version.

I removed all traces of previous version along with the SleepyHead App folder in MyDocuments prior to install of 0.8.8. First attempt to import data gave me a run time error failure at 99%. Repeat Import attempt was successful without any further issues.

For those newbies trying this...when importing you only navigate to the drive that the SD card is in. If you try to do the P Series folder you have gone too far. If you go too far you will likely get "can't find data" message.

Now to figure out what to do with all the extra data.

Edit to add image of what you see when you import and navigate to the SD card.
My SD card drive is E drive.... so all you need is the drive letter like this image and then click "choose"

That crash isn't good. :-/

Also, I thought I fixed that SD selection confusion once and for all. Oh well..

I'm not sure of the accuracy of the new graphs.. They are a little rough.. But there is only a very basic 5Hz flow signal to start with, so not a lot more I could do..
I did check them overlayed against an ASV machines data, and they followed the trends ok.. Tidal volume may still needs a little work to get the scale right..

Now that this is the morning after I updated, I have found that the import dialog does not allow me to import data directly from the card. The import dialog does not show mounted volumes and, like Finder, hides hidden directories, including "/Volumes", the parent of the mountpoints. Also, the newly derived graphs were flat again with the data from this morning's import. I got it to display after wiping my profile again.

tetragon wrote:Now that this is the morning after I updated, I have found that the import dialog does not allow me to import data directly from the card. The import dialog does not show mounted volumes and, like Finder, hides hidden directories, including "/Volumes", the parent of the mountpoints. Also, the newly derived graphs were flat again with the data from this morning's import. I got it to display after wiping my profile again.

I tried to do it this way to enable multiple folder selection, but it clearly doesn't work well enough.. I'm gonna have to drop the multiple selection and revert to the system native file dialog on windows platform.. It's rather clunky on Linux too..

It appears this version unfortunately really needs that "SleepApp" folder wipe..
If it's playing up during a subsequent import after starting fresh, then you have found a bug.. :/

jedimark wrote: If it's playing up during a subsequent import after starting fresh, then you have found a bug.. :/

I started fresh last night, after the rebuild. This morning had the second data import to the profile, and it was the second import data that lacked derived values. After I wiped again, this time by moving the data, it all came up again in a fresh, first import.

tetragon wrote:

jedimark wrote: If it's playing up during a subsequent import after starting fresh, then you have found a bug.. :/

I started fresh last night, after the rebuild. This morning had the second data import to the profile, and it was the second import data that lacked derived values. After I wiped again, this time by moving the data, it all came up again in a fresh, first import.

It might be the MultiThreading stuff playing up..

Could you rename your SleepApp folder, start a fresh one, then go into preferences and turn off "Enable Multithreading" in General, then try to reimport, and see if it breaks the same way?

Btw, what O/S is your computer running?

jedimark wrote: Could you rename your SleepApp folder, start a fresh one, then go into preferences and turn off "Enable Multithreading" in General, then try to reimport, and see if it breaks the same way?

With multithreading disabled, the derived graphs on the second import have non-zero calculated values. However, in this case, the AHI and snore graphs do not reliably render. On some days, all that appears is the title and y-tics.

I am still on OS X 10.6, Qt 4.7.4, running my own build from git.

tetragon wrote:

jedimark wrote: Could you rename your SleepApp folder, start a fresh one, then go into preferences and turn off "Enable Multithreading" in General, then try to reimport, and see if it breaks the same way?

With multithreading disabled, the derived graphs on the second import have non-zero calculated values. However, in this case, the AHI and snore graphs do not reliably render. On some days, all that appears is the title and y-tics.

I am still on OS X 10.6, Qt 4.7.4, running my own build from git.

I managed to reproduce the import crash with a test data set..

I just committed a fix to GIT that should resolve this.. The multithreaded code for once was not the culprit.. I did something silly by overrunning the buffer finding the last breath..

If it's still not working, I wouldn't mind a peek at your data to find out why..

If this fixes it, I'll update the windows build.. I've fixed a few other minor annoyances since releasing it..

jedimark wrote:

tetragon wrote:

jedimark wrote: Could you rename your SleepApp folder, start a fresh one, then go into preferences and turn off "Enable Multithreading" in General, then try to reimport, and see if it breaks the same way?

With multithreading disabled, the derived graphs on the second import have non-zero calculated values. However, in this case, the AHI and snore graphs do not reliably render. On some days, all that appears is the title and y-tics.

I am still on OS X 10.6, Qt 4.7.4, running my own build from git.

I managed to reproduce the import crash with a test data set..

I just committed a fix to GIT that should resolve this.. The multithreaded code for once was not the culprit.. I did something silly by overrunning the buffer finding the last breath..

If it's still not working, I wouldn't mind a peek at your data to find out why..

If this fixes it, I'll update the windows build.. I've fixed a few other minor annoyances since releasing it..

My graphs now display nicely with multithreading enabled.

I'm still having the missing graph problem with multithreading disabled on a build of 4bb339eec86d9affe5cf7304ea55235cf600f947. Does the number 27641229741079396352.00 mean anything to you (and do you know how much easier this message would be to type if I could copy data out of the sidebar)? It's the listed max for all of the missing graphs in the sidebar; the other numbers for the summary of the data look good. All affected nights are multi-session, but not all multi-session nights are affected. The missing graphs that I have seen are some combination of AHI, Leak, and Snore.

Anyway, I have data heading out your way now.

Gidday, Mark! Love the new version - you have obviously put a lot of effort into this since the last version. It is going to take a while to correlate what I see on the charts with how I feel. Maybe on the overview include a chart of how the slider is set - if the patient (especially me!) remembers to use that on a daily basis.

On the daily graph we now have a chart for AHI - is that a running cumulative score for the AHI over a shorter period than the full night? I know I have to work out what it means for me, but it is interesting to speculate.

Also, on the respiration rate chart on the overview page, there are several days when my rate drops to zero - would it be possible to step over missing data rather than have the graph drop to zero? Just makes for more pleasant reading! I know that I am still alive right now, so the zero obviously wasn't really zero ...!!!

Definitely more information to challenge the sleep doc with!

SleepyToo2 wrote:Gidday, Mark! Love the new version - you have obviously put a lot of effort into this since the last version. It is going to take a while to correlate what I see on the charts with how I feel. Maybe on the overview include a chart of how the slider is set - if the patient (especially me!) remembers to use that on a daily basis.

Slider? I'm not getting what you mean.

On the daily graph we now have a chart for AHI - is that a running cumulative score for the AHI over a shorter period than the full night? I know I have to work out what it means for me, but it is interesting to speculate.

That graph just shows the number of AHI events in an hour long window, stepped forward by 30 second increments. It helps identify event heavy areas..

Also, on the respiration rate chart on the overview page, there are several days when my rate drops to zero - would it be possible to step over missing data rather than have the graph drop to zero? Just makes for more pleasant reading! I know that I am still alive right now, so the zero obviously wasn't really zero ...!!!

This was a bug I just fixed.. I'm going to upload a new version tomorrow, as there was some crash issues as well that were fixed.. (plus a glitch with BMI calculations)

Definitely more information to challenge the sleep doc with!

They will probably laugh.. I haven't met a doctor who's really interested in data yet.
My Doc did at least remembered the name of my software.. That was rather cool..

tetragon wrote:Does the number 27641229741079396352.00 mean anything to you (and do you know how much easier this message would be to type if I could copy data out of the sidebar)? It's the listed max for all of the missing graphs in the sidebar; the other numbers for the summary of the data look good. All affected nights are multi-session, but not all multi-session nights are affected. The missing graphs that I have seen are some combination of AHI, Leak, and Snore.

Anyway, I have data heading out your way now.

That will be from an unitialized variable, or a nice buffer smash.. Foggy programmers should be locked safely away from computers..

I need to add a reindexer for CPAP data, to regenerate the summary data when things go haywire..

Second morning import without any issues.

AHI graph may cause some confusion. It appears to be just an hourly event number (OA, CA, Hyponeas)?? Similar to ResScan hourly Apnea event graph?? It sometimes matches the number of events during a specific hour and sometimes it comes close (at least from the quick looks I have seen).

Edit: I see you already addressed the AHI on graph question. I am good with it.

Example...last night during my peak AHI (per the graph) it showed hitting the 4 line and during the time frame of the "4" line....my total events was indeed 4.

Night before....there was a time frame showing total of 4 on the graph but there were only 3 that I could find.

During the hours with zero events of any kind the AHI graph line is on zero.

jedimark wrote:Slider? I'm not getting what you mean.

On the notes page, you have a slider that goes from "Zombie" to "Awesome"

Also, on the respiration rate chart on the overview page, there are several days when my rate drops to zero - would it be possible to step over missing data rather than have the graph drop to zero? Just makes for more pleasant reading! I know that I am still alive right now, so the zero obviously wasn't really zero ...!!!

jedimark wrote:This was a bug I just fixed.. I'm going to upload a new version tomorrow, as there was some crash issues as well that were fixed.. (plus a glitch with BMI calculations)

Will the program automatically update yet? I removed everything I could find on my hard disk for the old program (I had somehow got one directory buried within another, and I felt that was wrong - anyway, apart from the security on my PC (Win 7 Pro) the install of 0.8.8 went smoothly.

Definitely more information to challenge the sleep doc with!

jedimark wrote:They will probably laugh.. I haven't met a doctor who's really interested in data yet.
My Doc did at least remembered the name of my software.. That was rather cool..

Yeah, I know what you mean! My doc's eyes glazed over a bit when he saw the report that I prepared for him. I am sure the more detailed daily reports will really faze him... I hope that those of us who are really interested in our data will force at least some of the sleep docs into bringing their A game to our appointments, and to really give us some useful feedback. However, I recognize that the educational road is long and tortuous!

By the way, I like the "unintentional" leak graph - I know that is based on the total leak rate graph minus the minimum leak rate, but it still gives a useful indication of when we are having problems. Only challenge is I don't know whether it is me opening my mouth or if it is the mask becoming detached from my nostrils when I move during sleep. I do know that you are not going to be able to distinguish the difference without a video camera and other gear.

Oh, and can you add units to the y axes of the graphs?
((Edited to correct quotes))

SleepyToo2 wrote:

jedimark wrote:Slider? I'm not getting what you mean.

On the notes page, you have a slider that goes from "Zombie" to "Awesome"

Ah.. Thankyou.. I get what you mean now.. It should already be there?? If not it was one of the bugs I just fixed.. Tomorrow's build should fix it..

Will the program automatically update yet? I removed everything I could find on my hard disk for the old program (I had somehow got one directory buried within another, and I felt that was wrong - anyway, apart from the security on my PC (Win 7 Pro) the install of 0.8.8 went smoothly.

No auto updates yet.. Not far off.. and the frameworks done.. It's only been fog holding me back..

Yeah, I know what you mean! My doc's eyes glazed over a bit when he saw the report that I prepared for him. I am sure the more detailed daily reports will really faze him... I hope that those of us who are really interested in our data will force at least some of the sleep docs into bringing their A game to our appointments, and to really give us some useful feedback. However, I recognize that the educational road is long and tortuous!

Yup!

By the way, I like the "unintentional" leak graph - I know that is based on the total leak rate graph minus the minimum leak rate, but it still gives a useful indication of when we are having problems. Only challenge is I don't know whether it is me opening my mouth or if it is the mask becoming detached from my nostrils when I move during sleep. I do know that you are not going to be able to distinguish the difference without a video camera and other gear.

Motion sensing IR video cameras is getting a little fancy..
A simple microphone may be all that's needed, leaks have a very distinct sound, as does mouth breathing.. My wife hears when my mask leaks from across the room (I sleep on a recliner to avoid GERD killing me the next day.)

Oh, and can you add units to the y axes of the graphs?

Hover over the HTML panel and it will show them.. I'm was planning on adding tooltips showing it to the graph area you mentioned too. There isn't a lot of room to display them otherwise, without hogging valuable screen space. :/