🪑
CryoBench
  • CryoBench Manual
    • 📉Calculating FSC Metrics
    • 🤏Calculating Pose Errors
    • 🗺️Visualizing UMAP Clusterings of Latent Labels
  • Getting Started
    • 🛠️Installation Instructions
    • 🔭Running Reconstruction Models
Powered by GitBook
On this page
  1. CryoBench Manual

Visualizing UMAP Clusterings of Latent Labels

PreviousCalculating Pose ErrorsNextInstallation Instructions

Last updated 3 months ago

To plot cryoDRGN latent space UMAP clusterings colored by CryoBench ground truth labels, first run cryodrgn analyze with the current version of cryoDRGN to get UMAP results, and then load the relevant CryoBench environment to run the plotting script:

$ conda activate cryodrgn_model
(cryodrgn_model)$ cryodrgn analyze cryobench_input/003_IgG-1D_cdrgn2/ 19

$ conda activate cryodrgn_bench
(cryodrgn_bench)$ python ~/dev/CryoBench/metrics/visualization/visualize_umap_IgG-1D.py cryobench_input/003_IgG-1D_cdrgn2/ -o umap-plots/ --num_imgs 1000

🗺️