Maybe not the most important piece of data, but I'm wondering what might cause such an omission. Bug in software? Problem in data?
Anybody notice similar?
Thanks.
Hello Guest, Welcome to Apnea Board !
As a guest, you are limited to certain areas of the board and there are some features you can't use.
To post a message, you must create a free account using a valid email address.
Missing Leak Median

06272013, 08:42 AM
I've got all of 9days data downloaded into ResScan 4.3.0, but only two days (last night and two nights before that) show a "median" value on the "Leak" SummaryGraph. All the other days show the Leak median as 0 in text, and not at all on the graph. The other leak data Maximum and 95th percentile  is shown graphically and in text for all the days.
Maybe not the most important piece of data, but I'm wondering what might cause such an omission. Bug in software? Problem in data? Anybody notice similar? Thanks.
06272013, 01:29 PM
Use the SleepyHead, Luke.
"With ordinary talent and extraordinary perseverance, all things are attainable."  Thomas Foxwell Buxton
06272013, 04:34 PM
this how understood the question, forgive me if its not the correct answer
summary graph leak column shows leak value in colors median ... dark green  95% ... light green  maximum ... white median leak (dark green) is not shown when median leak is 0.0
06272013, 05:18 PM
Median is different than Average. My math might be off a little, but a simplified and unlikely scenario would be: If you use the machine for 8 hours and for 3 hours and 59 minutes you have a steady unintentional leak of 20Lm, but no leaks for the other 4 hours and 1 minute, then your average leak will be 9.xLm (almost 10Lm), but Median will be 0Lm because you spent so much time at 0Lm. You get the idea. Somebody who is better with math can correct me if that's wrong, but I get a Median of 0 almost every night and that's what I figure is going on. Could it be that ResScan is saying the Median is 0Lm on those nights because you spent a signifigant amount of time with no unintentional leaks on those nights? A quick look at the leak graph would shed some light.
06272013, 05:59 PM
Nope, you got it right. The average of a set of readings is taking the sum of the readings and dividing by the number of readings. The median of a set of readings is taking the readings, sorting them, and taking the middle value.
Median values are useful in the presence of reading glitches. For example, assume a sample set of 10, 11, 9, 12, 9, 9, 10000. The average is 151, but the median is 10. That one glitch reading is taking the average well away from what most of the data is indicating. However, the median value tends to ignore glitches.
06272013, 07:10 PM
(06272013, 04:34 PM)zonk Wrote: this how understood the question, forgive me if its not the correct answer I think you understood me correctly. All but two of the graphs had no darkgreen (median) portion, just the lightgreen (95th percentile), and a white (maximum). I guess some of the responders are saying that the median could actually have been zero on the days the data reported that. Thanks to all who responded, but I just don't get it. As I recall, median is the value that half the readings are above, and half are below. Unless the readings were zero all night long (which they were not) I would expect the median to always be some positive value, as it was for only two of the days I recorded.
06272013, 07:36 PM
The readings only have to be zero for just over half the night for the median value to be zero.
06272013, 08:20 PM
median leak = 50% percentile leak = at or below for 50% of the time using the machine
mr smiley face shows up when 70% percentile leak is less than 24 L/min resscan leak graph shows a red line at 24 L/min its good to be below the red line, above the line for brief period is not a problem
06282013, 08:17 AM
(06272013, 07:36 PM)RonWessels Wrote: The readings only have to be zero for just over half the night for the median value to be zero. Bingo! That comment did the trick for me. Zero is absolutely a reasonable value for the median of most of my data. Thanks! _____________ Brings up a second question: Since all masks are made with holes intended to "leak" to flush out exhaled air, why are there ever zeroes on the Leak graph? _____ PS: thought I posted similar last night, but don't see it. Hope it doesn't appear twice.
06282013, 08:41 PM
The venting is considered intentional leak. The machine and ResScan subtract intentional leak from total leak and report unintentional leak.

« Next Oldest  Next Newest »

Possibly Related Threads...  
Thread  Author  Replies  Views  Last Post  
Help Understanding Leak Rate & Statistics  Hojo  2  67 
Yesterday, 05:29 PM Last Post: Walla Walla 

Make Leak Advice/Help  JonnyCPAP  8  210 
09122017, 02:36 AM Last Post: JonnyCPAP 

[Equipment] NP10 exhaust leak.  stephen D  3  158 
09062017, 06:09 AM Last Post: silversnore 

Question about leak rate  Walla Walla  4  218 
09012017, 07:44 PM Last Post: Walla Walla 

FFM Cushion Leak with Data  Hojo  11  258 
08302017, 08:14 AM Last Post: bonjour 

[Equipment] Humidifier leak  dianewoodard53  3  213 
08272017, 05:38 PM Last Post: SarcasticDave94 

Where's the leak???  Jeffo1  3  223 
08212017, 06:22 PM Last Post: PaulaO2 