search

ahof1704 / VocalMat

Analysis of ultrasonic vocalizations from mice using computer vision and machine learning
https://www.dietrich-lab.org/vocalmat
Apache License 2.0
22 stars 10 forks source link

Pairwise distance matrices and their comparison #11

Closed hinada closed 2 years ago

hinada commented 2 years ago

Hi, I'd like to make figures of Pairwise distance matrices and their comparison in Fig 5B and Fig 5C in the eLife paper, at p10. I have identified variables needed. How can I make those figures?

ahof1704 commented 2 years ago

Hi,

The distance matrices are given as the output of the kernel alignment analysis. More specifically, you will find the variables associated with it at the end of the script: https://github.com/ahof1704/VocalMat/blob/9384fabfc1fbd9bc0ef8ca460b652e72c5b6819f/vocalmat_analysis/kernel_alignment.m#L1302)

Please note that kernel alignment analysis is marked as optional in the demo script. Thus, make sure you uncomment that section in order for the analysis to take place.

Good luck! Antonio

hinada commented 2 years ago

Hi Antonio,

Thank you for your reply. However, what I want to know is how to make matrix figures of Pairwise distance matrices and their comparison. Your comment is to show the diffusion map in 3D. In a default setting, matrix figures are neither shown nor saved. Do I have to make matrix figures manually using the variables you indicated? Is there any option to show and save matrix figures?

ahof1704 commented 2 years ago

Yes, the figures have to be made manually with the values stored in the Matlab variables. The figures could be plotted directly in Matlab, but the current version of the code does not offer that as an option. I would be glad to point out some resources to help you to code the heat map plot, if you'd like to do so.

Best Antonio

On Mon, Dec 13, 2021, 6:09 PM hinada @.***> wrote:

Hi Antonio,

Thank you for your reply. However, what I want to know is how to make matrix figures of Pairwise distance matrices and their comparison. Your comment is to show the diffusion map in 3D. In a default setting, matrix figures are neither shown nor saved. Do I have to make matrix figures manually using the variables you indicated? Is there any option to show and save matrix figures?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/ahof1704/VocalMat/issues/11#issuecomment-992994220, or unsubscribe https://github.com/notifications/unsubscribe-auth/AB52JRBSZCS44I5OGN23TDTUQZ4LFANCNFSM5JYLY2YQ . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

hinada commented 2 years ago

Hi Antonio,

Thank you for your reply. I guessed I can use the heatmap() function of MATLAB and valuables dist_X1 and dist_X2 in kernel_alignment.m to make matrix figures of Pairwise distance matrices for each group. For the preparation of comparison of Pairwise distance matrices, I guess I have to use the corrcoef() function and each dist_X[i] valuable one by one. Am I correct?

ahof1704 commented 2 years ago

Correct, you can use Matlab's heatmap function to obtain the pairwise distance plot (figure 5b). Then, for the similarity plot (5c), you should compute the correlation between the two matrices as a whole, not value by value. Such correlation is computed in the kernel_alignment function as shown below https://github.com/ahof1704/VocalMat/blob/9384fabfc1fbd9bc0ef8ca460b652e72c5b6819f/vocalmat_analysis/kernel_alignment.m#L766

hinada commented 2 years ago

Got it! Thank you so much!