Compression rate problem in Demo Compression & Denoising

[RenieWell]

Hi!
I have followed your excellent work in your preprint paper Penalized matrix decomposition for denoising, compression, and improved demixing of functional imaging data, but I have different result in compression rate.
The paper claimed Patch-wise PMD achieved a compression rate of 180 in voltage data. However, after I saved the compressed file of the voltage data, it came out to be a compression rate of 3 for measuring the file size before and after the compression. And I used the sample data you provided in https://github.com/ikinsella/trefide.
Could you help me out?