Bland-Altman Analysis of Grape 1 vs Grape 2

30 views
Skip to first unread message

Charles Skoning

unread,
Apr 12, 2024, 1:12:21 PMApr 12
to hamsci-grape
Hello all, 

I thought I would compare the data from my Grape 1 vs Grape 2 for a twenty four hour period to see how close they were to each other. I used a statistical method called Bland-Altman which is used to compare two methods of measuring the same thing, in this case, Doppler shift. Here are my observations:

For my particular station my Grape 1 has a slight bias of 0.002955 Hz relative to the Grape 2.

The margin of error between the Grapes is +/- 0.001001 Hz.

I've attached a scatter plot with 95% confidence levels displayed. 

Not sure if this is useful, but I thought I would give it a go and share my results. Both Grapes agree with each other very well in my opinion.

Kind regards, 
Chuck, KE9SA
Bland_Altman_G1_VS_G2.png

Jonathan

unread,
Apr 12, 2024, 1:23:29 PMApr 12
to Charles Skoning, hamsci-grape
Hi Chuck,

I suspect this is due to the error introduced by the soundcard's clock on the Grape 1 as the error is mHz. This is what would account for the difference.

Jonathan
KC3EEY 

--
You received this message because you are subscribed to the Google Groups "hamsci-grape" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hamsci-grape...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hamsci-grape/a8521300-ae0b-43ad-a4fc-162f9d175351n%40googlegroups.com.

Jonathan

unread,
Apr 13, 2024, 11:27:41 AMApr 13
to hamsci-grape
Hi Chuck,

I wanted to show you some plots that illustrate how a soundcard's sample rate can drift over the course of the day. There is a plot, specifically the bottom subplot, that shows the drift of sample rate throughout the day on the Audio Injector Stereo soundcard. This drift is corrected with the use of a pulse per second signal from a GPS fed into the other channel of the soundcard using vlfrx-tools software. I have proposed the use of vlfrx-tools and a GPS PPS as an alternative software for the Grape.
Screen Shot 2024-04-13 at 11.11.24 AM.png
With temperature variations throughout the day, sample rate can drift considerably enough to cause some error.

Jonathan

On Fri, Apr 12, 2024 at 1:36 PM Charles Skoning <charles...@gmail.com> wrote:

Jonathan,

Interesting, I did not consider that.

Chuck, KE9SA

Reply all
Reply to author
Forward
0 new messages