search

david1412 / Time-varying--MIMO

Time varying system identification of MIMO Acoustics systems
0 stars 0 forks source link

systemdistance extension #5

Closed david1412 closed 6 years ago

david1412 commented 7 years ago

I am trying out system distance with different angular speeds and i come up with the new update in mean time.

david1412 commented 7 years ago

system distance with angular speeds .. once check the plot sd im trying out 1. critical 2. over and under sampling with different angular speeds.

narahahn commented 7 years ago
david1412 commented 7 years ago

. I will modify it . Sigma* / its Omega there systemdistance

i labelled it with axes .. its for your reference .

narahahn commented 7 years ago

If you make changes, please push it to the repository so that I can test it.

david1412 commented 7 years ago

I used the desktop version and i forgot to push at that time.. now I pushed into repository .

david1412 commented 7 years ago

New plot: system identification new image

narahahn commented 7 years ago

I guess the angular speeds are not

0.5, 4, 8

but

2*np.pi/0.5, 2*np.pi/4, 2*np.pi/8
david1412 commented 7 years ago

I gave the angular speeds Q=[8,4,0.5]

narahahn commented 7 years ago

Please push all the changes to the repository, so that I can run the same code as yours.

david1412 commented 7 years ago

I pushed all my changes to the repository . please check

narahahn commented 7 years ago

The parameter Q determines the angular speed of the microphone Omega which is

Omega = 2 * np.pi / Q

In your recent script, one of the Q value is 4.57. I guess you considered the upper bound of the angular speed that avoid aliasing,

c / N / R

which is 4.573333... in your experimental setup. However, c / N / R should be Omega not the parameter Q. If you want to run a simulation in which the angular speed is exactly c / N / R, you should find a Q that satisfies

c / N / R = 2 * np.pi / Q
david1412 commented 7 years ago

yes i understood the concept ,

i took c=343 and computed with the formula . I got Q= 0.018318324510727655 should i take the values like Q=0.19 etc ?

david1412 commented 7 years ago

I am getting this error ,, TypeError: denominator() missing 1 required positional argument: 'psi'

in line 418 .. whats the error about ?

narahahn commented 7 years ago

According to its definition denominator requires 2 input arguments.

narahahn commented 7 years ago

The indentations are all messed up. How are you pushing your changes to the repository? Do you copy your script and paste it in the web browser?

Please try to use the push and pull functionality in git. This will avoid most of the problem we are having right now.

david1412 commented 7 years ago

I am using Github desktop version . I am opening the system distance file and making changes .

opening the file>data insertion>save>commit changes with description>push .. I am using push only ..

starting i used to copy the code in the git but from few weeks im using git desktop version .

david1412 commented 7 years ago

i am getting one error in the code : i changed the denominator to File "E:/Thesis/projects/new/Neuer Ordner/new11.py", line 419, in denom = denominator(impulse_response[:,psi])

h(phi,n) > orginal impulse response

TypeError: denominator() missing 1 required positional argument: 'psi' only one error im getting

narahahn commented 7 years ago

According to its definition, denominator() takes 2 input arguments, but only one is used. So you can change it such that it only takes 1 input that is actually used.

https://github.com/david1412/Time-varying--MIMO/blob/809a33b60f3a8dbd36597768efb9bc6c65335e82/systemdistance.py#L13-L14

Then you have to change the following line:

https://github.com/david1412/Time-varying--MIMO/blob/809a33b60f3a8dbd36597768efb9bc6c65335e82/systemdistance.py#L122