Skip to content

Instantly share code, notes, and snippets.

@rodonn
Last active October 30, 2023 21:10
Show Gist options
  • Save rodonn/29e3f9b249fa28f122f8f741fc5ea57e to your computer and use it in GitHub Desktop.
Save rodonn/29e3f9b249fa28f122f8f741fc5ea57e to your computer and use it in GitHub Desktop.
Python Parallel
from concurrent.futures import ThreadPoolExecutor # Or ProcessPoolExecutor for CPU-bound tasks
with ThreadPoolExecutor() as executor:
future1 = executor.submit(slow_func1, a, b, c)
future2 = executor.submit(slow_func2, x, a, z)
x = future1.result()
y = future2.result()
return x + y
from concurrent.futures import ThreadPoolExecutor
from concurrent.futures import Future
import pandas as pd
import random
class LazyDataFrame:
def __init__(self, query):
self.executor = ThreadPoolExecutor()
self.future = self.executor.submit(read_sql_query, query)
def __getattr__(self, name):
if hasattr(pd.DataFrame, name):
# Block until the future is resolved
df = self.future.result()
return getattr(df, name)
else:
raise AttributeError(f"'LazyDataFrame' object has no attribute '{name}'")
def result(self):
return self.future.result()
# Example usage:
from concurrent.futures import ThreadPoolExecutor
def read_sql_query(query):
import time
duration = random.randint(5, 10)
time.sleep(duration)
print(f"Finished query: {query} (duration: {duration}s)")
return pd.DataFrame({"data": [query]})
lazy_df1 = LazyDataFrame("query1")
lazy_df2 = LazyDataFrame("query2")
print("I can do stuff without waiting for the queries to finish")
print(lazy_df1.head())
print(lazy_df2.head())
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment