trends for astrophysics look exactly the same as trends for cogsci, s…

…uggesting theyre really just trends for computer science and that semanticscholar is not a great resource for isolating to a specific field
nathimel · Aug 1, 2023 · 96e2b3c · 96e2b3c
1 parent d3cf8ee
commit 96e2b3c
Show file tree

Hide file tree

Showing 4 changed files with 171 additions and 164 deletions.
diff --git a/src/examples/scratch/outputs/atlas_s2-7-25-23_hafenetal/README.md b/src/examples/scratch/outputs/atlas_s2-7-25-23_hafenetal/README.md
diff --git a/src/examples/scratch/outputs/atlas_s2-7-29-23_centered_imeletal/README.md b/src/examples/scratch/outputs/atlas_s2-7-29-23_centered_imeletal/README.md
@@ -8,3 +8,6 @@ First 41 pubs are ridiculously slow.
 Finished overnight with no issues, though. 
 And second run to 20k does not seem to be too slow either.
 Trying to expand to 30k, things seem to be slowing down. Maybe getting throttled.
+After a little bit things get faster.
+Some rounds of 500 calls take 2 mins, others take more than 20. Not sure why.
+Got to 30k.
diff --git a/src/examples/scratch/outputs/atlas_s2-7-30-23_centered_hafenetal/README.md b/src/examples/scratch/outputs/atlas_s2-7-30-23_centered_hafenetal/README.md
@@ -0,0 +1,5 @@
+# README.md
+
+30k pubs
+
+Bad news though. The trends look exactly like the trends for cogsci. This suggests to me that semantic scholar is actually not a good resource for _isolating_ a field of scientific inquiry. Because clearly what's happenning is that the majority of these papers are just in computer science, probably 90% AI.
diff --git a/src/examples/scratch/temporal.ipynb b/src/examples/scratch/temporal.ipynb