Revision—workflow updates with new math results

[murrayds]

Revised the workflow to produce the following:

• Add wildcard for setting gamma in embedding models between 0.75 and 1.0. All files building on the embeddings (most) are therefore marked with a gamma value
• Add wildcard for ways of measuring the size of institutions. ALl files building on these files are now marked with a size wildcard
  • all: count size using all unique mobile and nonmobile researchers at the organization
  • mobile: count size using all unique mobile researchers at the organization only
  • freq: count size of organizations using total frequency across trajectories, the same way frequency would be counted with natural language
• Plot gravity & predicted vs. actual results with new SVD cosine distance and Laplacian Eigenmap cosine distance.
• Added missing packages, R-viridis and tqdm, to the conda environment
• Minor bug-fixing to make sure the workflow runs

[jisungyoon]

Do you have a table of figures for this additional analysis?

[murrayds]

@jisungyoon see the slack channel