All about SleepAnalyser - Page 33 - maemo.org

Active Topics

The N900 is now 15 years old! (9)
to Nokia N900 by Arno_11 - 4 hrs, 58 mins ago
Sailfish OS for the Motorola Moto G7 Power (XT1955-5) - (ocean) (4)
to SailfishOS by Maemish - 5 days, 17 hrs ago
more...

Page 33 of 49

Thread Tools

x-lette	2011-01-27 , 11:55
Posts: 539 \| Thanked: 165 times \| Joined on Feb 2010 @ Berlin, Germany	#321

Originally Posted by caco3

This should be fixed in the recent version. If not, please let me know.

Will try that tonight, thanks!

I can implement to add a external application, that should be quite easy.
Is it possible to run EvilAlarm from command line and set off an alarm immediately? Could you have a look if that is possible?

I got so used to apps being controllable by commandline I even didn't think about an app with GUI only.

Sadly it's a binary executable so I'll have to contact the author to get more infos.

Quote & Reply |

	caco3	2011-01-27 , 14:37
	Posts: 560 \| Thanked: 423 times \| Joined on May 2010 @ Switzerland	#322

Originally Posted by x-lette

Another two suggestions:
Could you fit the trigger level to the other numberings? Actually the level is set to roughly 1/10 of the other values. Would be nice if you could suite the level to the other values so all of them are consistent.

Hi X-Lette

Could you explain a bit more, what you exactly suggest?
Why would you like to have the trigger value changed?
It does not really have much to do with the scaling values!
And how do you get the ratio 1/10?
If I have Y scaling on 40, and the trigger on approx 17, that makes only a ratio of 1/2.4

__________________
On N9 check out this:
CacheMe 4 the N9, a geocaching client / MiniBible, a bible viewer / TheWord brings daily bible verses onto your phone / BatteryGraph to monitor the battery drainage / doublepress2unlock to unlock your phone with a double press onto the power button / GPRS Data Usage to monitor your GPRS data usage /
and more...

On N900 check out this: SleepAnalyser to analyse your sleep movements / PasswordMaker a for a password generator

Quote & Reply |

x-lette	2011-01-27 , 15:09
Posts: 539 \| Thanked: 165 times \| Joined on Feb 2010 @ Berlin, Germany	#323

Originally Posted by caco3

Could you explain a bit more, what you exactly suggest?
Why would you like to have the trigger value changed?

It's just for consistence. When playing around with values for yellow and blue graph and then starting recording and trying to adjust the trigger level to some sensible value one still has in mind the much higher values of blue graph. I found it hard to catch the right trigger level that suites to blue graph and after all I ended up with trial-and-error and had to look at old values and how far they are away from what I think would be perfect.

It does not really have much to do with the scaling values!
And how do you get the ratio 1/10?
If I have Y scaling on 40, and the trigger on approx 17, that makes only a ratio of 1/2.4

I see the values doesn't have to be linear. I set blue graph to 60 (formerly 70) and trigger level is now on 7. That's why I thought they were roughly like 1:10. (I realize that I still didn't understand the concept of the values yet even if you already tried to explain)

Quote & Reply |

Mentalist Traceur	2011-01-27 , 15:48
Posts: 2,225 \| Thanked: 3,822 times \| Joined on Jun 2010 @ Florida	#324

BTW, sorry I haven't paid much attention to this thread:

I LOVE the switch to the blue instead of the yellow. Before I could get woken up by an errant movement when not nearly in the light sleep stages due to the alarm triggering by the 'yellow' accelerometer value. The blues do a decent job of approximating my sleep, at least.

Also, I noticed the new UI approach when you first implemented lucid dreaming / wake up stuff, and that was pretty neat.

Quote & Reply |

	caco3	2011-01-27 , 21:02
	Posts: 560 \| Thanked: 423 times \| Joined on May 2010 @ Switzerland	#325

Originally Posted by x-lette

I realize that I still didn't understand the concept of the values yet even if you already tried to explain

If you show me where you struggle, I can try to explain it.

Here just a summary:
Y scaling (yellow) only scales the visual indication of the yellow movement bars. No other usage.
Y scaling (blue) only scales the visual indication of the blue sleep pattern graph. No other usage.
Trigger level defines when the alarm gets set off. If the blue graph is over the trigger level, the alarm gets activated.

Now, the trigger level is connected to the blue graph (obviously). if you stretch the blue graph more, the trigger level also will be drawn higher. This means, it goes with the blue graph.
Changing the scaling of the blue graph does not change the sensitivity. Changing it does nothing but change the visualization. Because of that the trigger level will always have the same value.

Just have a look on the 2 attached screen shots. The first one has an Y scaling (blue) of 40, the 2nd one of 60.
On both the trigger level is on 23. As you can see, the alarm got set off when the blue sleep pattern graph crossed the trigger line.
On both screen shots, the trigger line has the same relative position to the blue sleep pattern graph.

Now, how to set it up properly:
Do a recording for a night, or use one of the already existing records.
Now adjust the Y scaling (blue), so it is most of the time in light or deep sleep. I would say, on the shown example 40 seems to be a good value, maybe 30 or 50 would work too. How ever more than 60 seems to be too much, as it indicates you have been awake very often. Unless you had a very rough night, it is not realistic.
After you set the blue graph scaling, you can set the trigger level on the next recording. I suggest it to have it on top or a bit above the light sleep area. The slightly gray background will help you for that set up.
Unless you change your bed or position of the phone, the scalings will not have to be changed anymore. Changing the trigger level should be enoughs for the fine adjustments.

Changing the Y scaling for the yellow bars is actually not needed anymore since we use the blue graph for triggering.

Attached Images

Last edited by caco3; 2011-01-27 at 21:08.

Quote & Reply |

The Following 2 Users Say Thank You to caco3 For This Useful Post:
Mentalist Traceur, x-lette

prattmic	2011-01-28 , 05:12
Posts: 29 \| Thanked: 2 times \| Joined on Apr 2010 @ NC, USA	#326

I tryed out the extras version of this and quite like it, so I decided to update to the latest version, but now I get an error when I try to open the program. I thought it was an issue with the config file, so I deleted ~/.SleepAnalyser/SleepAnalyser.conf, but that didn't help. I'm guessing that something else is wrong, or that was the wrong config file. Anyways, here is my log file, and thankd for the help and the great program!

Code:

Friday 01/28/11 00:06:30
SleepAnalyser
Version: 2.0-9
Locale: en_US , Encoding: ISO8859-1
-----------------------------------------------------------------------------

## An error occurred!
###################### RECORD DATA #######################
Index: -1
LastIndex: -1
StartTime_tuple: -1
StartTime_seconds: -1
EndTime_tuple: -1
EndTime_seconds: -1
AlarmIsSet: False
AlarmTime_tuple: -1
AlarmTime_seconds: -1
AlarmTimeWindowStart_tuple: -1
AlarmTimeWindowStart_seconds: -1
AlarmTimeWindowEnd_tuple: -1
AlarmTimeWindowEnd_seconds -1
AlarmTimeWindow_minutes: -1
Note: 
Sampling_interval: -1
Sample_smoothing: -1
Value_smoothing: -1
Trigger: -1
Status: -1 
Use_Lucid_Dream_Song: False
Lucid_Dream_Song_Start_tuple: -1
Lucid_Dream_Song: 
Lucid_Dream_Song_Delay: -1
Lucid_Dream_Song_Duration: -1
Lucid_Dream_Song_Volume: -1
Use_WakeUp_Song: False
WakeUp_Song_Start_tuple: -1
WakeUp_Song: 
WakeUp_Song_Duration: -1
WakeUp_Song_Volume: -1
##########################################################
###################### CONFIG DATA #######################
('yscale_sleep_pattern', '40')
('lucid_dream_song_duration', '0')
('sampling', '1000')
('limitmovement', '250')
('user', '')
('use_wakeup_song', 'False')
('wakeup_song', '')
('derivation_offset', '2')
('stopafteralarm', 'True')
('wakeup_song_duration', '0')
('lucid_dream_song_volume', '30')
('alarmtimewindow', '30')
('delaystarttime', '0')
('alarmhour', '7')
('derivation', '5')
('trigger', '14')
('sample_smoothing', '10')
('silentprofile', 'False')
('ysize', '121')
('wakeup_song_volume', '50')
('lastrecord', '')
('alarmminute', '0')
('usealarm', 'False')
('value_smoothing', '10')
('yoffset_sleep_pattern', '0')
('xscale', '1')
('lucid_dream_song', '')
('offlinemode', 'False')
('yscale_movement', '40')
('lucid_dream_song_delay', '600')
##########################################################
Traceback (most recent call last):
  File "/opt/SleepAnalyser/main.py", line 2683, in <module>
    main() # Now we can start it.
  File "/opt/SleepAnalyser/main.py", line 138, in main
    config=LoadConfig()
  File "/home/builder3/maemo-fremantle-armel-extras-devel/work/sleepanalyser-2.0/debian/sleepanalyser/opt/SleepAnalyser/functions_general.py", line 234, in LoadConfig
    QMessageBox.critical(None, "Changes in SleepAnalyser",
NameError: global name 'QMessageBox' is not defined

Quote & Reply |

	caco3	2011-01-28 , 07:42
	Posts: 560 \| Thanked: 423 times \| Joined on May 2010 @ Switzerland	#327

hmm...
it seems that the autobuilder once again created a corrupted package. I will reupload it this evening.

Quote & Reply |

The Following User Says Thank You to caco3 For This Useful Post:
Mentalist Traceur

x-lette	2011-01-28 , 09:30
Posts: 539 \| Thanked: 165 times \| Joined on Feb 2010 @ Berlin, Germany	#328

Originally Posted by x-lette

Sadly it's a binary executable so I'll have to contact the author to get more infos.

EvilAlarm does have a commandline switch to immediately start getting on your nerves:

Code:

evilalarm --wakeup

does the trick. Seems like there are still some bugs with sound output but at least we could combine these two.

Quote & Reply |

x-lette	2011-01-28 , 10:04
Posts: 539 \| Thanked: 165 times \| Joined on Feb 2010 @ Berlin, Germany	#329

Originally Posted by caco3

If you show me where you struggle, I can try to explain it.

I guess I now understood the relations ...

Now, the trigger level is connected to the blue graph (obviously). if you stretch the blue graph more, the trigger level also will be drawn higher. This means, it goes with the blue graph.

That's one point which wasn't quite obvious to me. I changed scaling and trigger level line also changed so I thought the trigger level itself did change in any way.

Changing the scaling of the blue graph does not change the sensitivity. Changing it does nothing but change the visualization. Because of that the trigger level will always have the same value.

Yes, the value stays the same but it's a bit unusual that the scaling of the line changes without having changed the corresponding value.

Now, how to set it up properly: [...]

That's quite the way I did try to setup working values but I still have to adjust them. But now I know that the trigger level is only drawn at a different height when changing blue value but still triggers at same movement level.

Thanks for your explanations (again)!

Quote & Reply |

Joseph9560	2011-01-28 , 11:49
Posts: 440 \| Thanked: 160 times \| Joined on Aug 2010 @ Las Vegas, NV	#330

Nice app! Before trying this app, I thought it would be based on microphone sound recording. I think there is a possibility of adding one more functionality to make it something like happy wake up app for symbian. Just start recording only some time before the alarm time, then activate alarm if the use is almost awake else no alarm.