How to adjust the code for other datasets

[atheeraa]

Hello, Thank you for sharing the code.

I need help understanding how to adjust the Pseudo_endmembers_generation.m for other datasets. From the other issues I see you said that k is the number of patches, but it affects the resulting dimensions Also, in this line: [sub_EM, ind, ~] = VCA(sub_X_2d, 'Endmembers', 5, 'SNR', 30);

What does the 5 represent? Do I set it to be a larger number than the actual number so that the algorithm finds the actual number? If I set to the actual number of endmembers, EM and Abund dimensions are not correct,

My dataset is as the following: X = 156 9025 (Number of bands x number of pixels) M = 156 3 (Number of bands x number of endmembers) Abundance = 9025 * 3 (Number of pixels x number of endmembers

Now , what I get if I set k = 5 and set this line as: [sub_EM, ind, ~] = VCA(sub_X_2d, 'Endmembers', 3, 'SNR', 30);

I get :

Abund= 3 1083 EM = 156 1083

[danfenghong]

The k means how many patches you would like to separate the image spatially. If you set k to 5, that means the patch size should be 5*5.

Atheer Abdullah @.***> 于2023年11月19日周日 20:42写道：

Hello, Thank you for sharing the code.

I need help understanding how to adjust the Pseudo_endmembers_generation.m for other datasets. From the other issues I see you said that k is the number of patches, but it affects the resulting dimensions Also, in this line: [sub_EM, ind, ~] = VCA(sub_X_2d, 'Endmembers', 5, 'SNR', 30);

What does the 5 represent? Do I set it to be a larger number than the actual number so that the algorithm finds the actual number? If I set to the actual number of endmembers, EM and Abund dimensions are not correct,

My dataset is as the following: X = 156 9025 (Number of bands x number of pixels) M = 156 3 (Number of bands x number of endmembers) Abundance = 9025 * 3 (Number of pixels x number of endmembers

Now , what I get if I set k = 5 and set this line as: [sub_EM, ind, ~] = VCA(sub_X_2d, 'Endmembers', 3, 'SNR', 30);

I get :

Abund= 3 1083 EM = 156 1083

— Reply to this email directly, view it on GitHub https://github.com/danfenghong/IEEE_TNNLS_EGU-Net/issues/13, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFL2GZXL4QXSFPRWPCKL2B3YFH5FDAVCNFSM6AAAAAA7RWYAGOVHI2DSMVQWIX3LMV43ASLTON2WKOZSGAYDAOBTGQ4TMMI . You are receiving this because you are subscribed to this thread.Message ID: @.***>

[atheeraa]

Thank you for replying, what about the number of endmembers in the line: [sub_EM, ind, ~] = VCA(sub_X_2d, 'Endmembers', 3, 'SNR', 30);

[danfenghong]

The number of endmemebers can be determined by algorithms.

Atheer Abdullah @.***> 于2023年11月19日周日 22:06写道：

Thank you for replying, what about the number of endmembers in the line: [sub_EM, ind, ~] = VCA(sub_X_2d, 'Endmembers', 3, 'SNR', 30);

— Reply to this email directly, view it on GitHub https://github.com/danfenghong/IEEE_TNNLS_EGU-Net/issues/13#issuecomment-1817865757, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFL2GZQKORDH3KJ6O7OEW2TYFIG5DAVCNFSM6AAAAAA7RWYAGOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQMJXHA3DKNZVG4 . You are receiving this because you commented.Message ID: @.***>

[atheeraa]

Which is EM? EM's dimensions are 156*1805

I don't get it. I know 156 is the number of the bands, but what is the number of the endmembers?

[danfenghong]

156 is spectral band rather than endmembers.

Atheer Abdullah @.***> 于2023年11月19日周日 22:30写道：

Which is EM? EM's dimensions are 156*1805

I don't get it. I know 156 is the number of the bands, but what is the number of the endmembers?

— Reply to this email directly, view it on GitHub https://github.com/danfenghong/IEEE_TNNLS_EGU-Net/issues/13#issuecomment-1817872137, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFL2GZTJHUWXI2W7YRVGMYDYFIJYTAVCNFSM6AAAAAA7RWYAGOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQMJXHA3TEMJTG4 . You are receiving this because you commented.Message ID: @.***>

[atheeraa]

Yes , what about 1805? it's not the # of endmembers, nor any dimension of the data

and how can i know the number of endmembers extracted by the algorithm? Where is it? which variable?