import pandas as pd

repo = "Chainticks/labor-demand-index"
latest = pd.read_csv(f"https://huggingface.co/datasets/{repo}/resolve/main/LATEST_DATE.txt", header=None).iloc[0, 0]

# The combined index: where automatable work is in demand AND hard to fill.
url = f"https://huggingface.co/datasets/{repo}/resolve/main/labor_pressure_index/date={latest}/part-0000.parquet"
df = pd.read_parquet(url)
print(df.sort_values("labor_pressure_score_0_1", ascending=False)[
    ["role_family", "geo", "labor_pressure_score_0_1", "blended_vacancy_hire_ratio", "stale_share_30d", "salary_mid_median"]
].head(10))