Select optimized results and transition states

Export stable optimized structures and transition-state candidates separately.

Python

Python

from molop import AutoParser

batch = AutoParser("water_mp2.out", n_jobs=1)
normal = batch.filter_state("normal")
optimized = normal.filter_state("opt")
transition_states = normal.filter_state("ts")

print({
    "parsed": len(batch),
    "normal": len(normal),
    "optimized": len(optimized),
    "transition_states": len(transition_states),
})

optimized.to_summary_df(
    brief=False, flatten_columns=True
).to_csv("optimized.csv", index=False)

transition_states.to_summary_df(
    brief=False, flatten_columns=True
).to_csv("transition_states.csv", index=False)

Output

Selection counts

Text Only

{'parsed': 1, 'normal': 1, 'optimized': 0, 'transition_states': 0}

The generated files are:

Created files

Text Only

optimized.csv
transition_states.csv

The counts above are for the bundled single-point sample; counts and whether the two CSV files have data depend on the input set. Use Vibration.num_imaginary in the transition-state CSV to review imaginary-mode counts. A missing column means selected frames had no structured frequency result.

Replace water_mp2.out with results/*.log when selecting optimization or transition-state jobs.

CLI

Bash

molop -q parse "results/*.log" \
  filter-state --state normal \
  filter-state --state ts \
  to-summary-df --full --out transition_states.csv

Created file

Text Only

transition_states.csv

Scientific review

is_TS is a programmatic candidate decision, not a replacement for checking the imaginary mode, connectivity change, and reaction path. Database ingestion and automated reaction workflows should review the displacement and endpoint chemistry.