如何将public_client返回的Generator转为Pandas DataFrame并解决报错

Yep, that stray 'message' string at the end of your list is exactly what's causing the AttributeError! Here's why and how to fix it:

Why the Error Happens #

pd.DataFrame.from_records() expects every element in your input to be a dictionary (or an object with a .keys() method) so it can map the keys to DataFrame columns. That random string 'message' doesn't have a .keys() attribute—hence the error you're seeing.

Fix: Filter Out Non-Dictionary Elements #

You can solve this by filtering out any elements in the generator that aren't dictionaries before converting to a DataFrame. This is efficient to do directly with the generator (no need to convert it to a list first, which saves memory):

import pandas as pd

# Filter the generator to keep only dictionary items
filtered_trades = (item for item in data if isinstance(item, dict))
df = pd.DataFrame.from_records(filtered_trades)

If you prefer to first convert the generator to a list (to inspect it easily), you can filter the list instead:

import pandas as pd

# Convert generator to list first
trade_list = list(data)
# Filter out non-dictionary elements
filtered_trades = [item for item in trade_list if isinstance(item, dict)]
df = pd.DataFrame(filtered_trades)

Bonus: Inspect the Mysterious 'message' #

If you want to check what that 'message' string contains (it might be a warning or error from the API), you can extract it separately:

trade_list = list(data)
# Split into valid trade dictionaries and other messages
valid_trades = [item for item in trade_list if isinstance(item, dict)]
api_messages = [item for item in trade_list if not isinstance(item, dict)]

# Print the messages to debug
print("API Messages:", api_messages)

# Create DataFrame from valid trades
df = pd.DataFrame(valid_trades)

内容的提问来源于stack exchange，提问作者spotnag