[Python] Polars: What is the way to limit number of threads? [duplicate]

Polars is consuming all the threads apparently in my app deployed using gunicorn+supervisor. What is the best way to limit the number of threads for polars? I am not reading a csv or writing a csv (using in below code for example only)? also, any way to log the number of threads being consumed by polars at any time?.

├── polars_script.py

import polars as pl

df = pl.read_csv("input.csv")

# Perform some basic operations
df = df.filter(pl.col("column_name") > 50) # Filter rows where 'column_name' > 50
df = df.with_column((pl.col("column_name") * 2).alias("new_column")) # Add a new column

# Write the DataFrame to a new CSV file
df.write_csv("output.csv")

print("Data processing complete. Output saved to output.csv")


├── app.py

from flask import Flask
import polars as pl

app = Flask(__name__)

@app.route('/process')
def process_data():
# Read a CSV file into a Polars DataFrame
df = pl.read_csv("input.csv")

# Perform some basic operations
df = df.filter(pl.col("column_name") > 50) # Filter rows where 'column_name' > 50
df = df.with_column((pl.col("column_name") * 2).alias("new_column")) # Add a new column

# Write the DataFrame to a new CSV file
df.write_csv("output.csv")

return "Data processing complete. Output saved to output.csv"

if __name__ == "__main__":
app.run()


├── gunicorn_config.py

bind = "0.0.0.0:8000"
workers = 20



gunicorn -c gunicorn_config.py app:app

Continue reading...