Skip to content

Chunking data inserted into knowledge base #195

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
ea-rus merged 1 commit into from
May 12, 2025
Merged

Chunking data inserted into knowledge base #195

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
32 changes: 27 additions & 5 deletions mindsdb_sdk/knowledge_bases.py
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
import copy
import json
from typing import Union, List
from typing import Union, List, Iterable

import pandas as pd

Expand All @@ -16,6 +16,19 @@
from .query import Query
from .databases import Database

MAX_INSERT_SIZE = 1000


def split_data(data: Union[pd.DataFrame, list], partition_size: int) -> Iterable:
"""
Split data into chunks with partition_size and yield them out
"""
num = 0
while num * partition_size < len(data):
# create results with partition
yield data[num * partition_size: (num + 1) * partition_size]
num += 1
Comment on lines +22 to +30
Copy link
Copy Markdown

@entelligence-ai-pr-reviews entelligence-ai-pr-reviews bot May 12, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Correctness: The split_data function doesn't handle empty datasets, which could lead to an infinite loop if an empty DataFrame or list is passed to it.

📝 Committable Code Suggestion

‼️ Ensure you review the code suggestion before committing it to the branch. Make sure it replaces the highlighted code, contains no missing lines, and has no issues with indentation.

Suggested change
def split_data(data: Union[pd.DataFrame, list], partition_size: int) -> Iterable:
"""
Split data into chunks with partition_size and yield them out
"""
num = 0
while num * partition_size < len(data):
# create results with partition
yield data[num * partition_size: (num + 1) * partition_size]
num += 1
def split_data(data: Union[pd.DataFrame, list], partition_size: int) -> Iterable:
"""
Split data into chunks with partition_size and yield them out
"""
if len(data) == 0:
return
num = 0
while num * partition_size < len(data):
# create results with partition
yield data[num * partition_size: (num + 1) * partition_size]
num += 1

Copy link
Copy Markdown

@fshabashev fshabashev May 12, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

AI is wrong here because if data is empty then len(data) is 0 and the loop body won't be executed



class KnowledgeBase(Query):
"""
Expand Down Expand Up @@ -152,7 +165,7 @@ def insert_webpages(self, urls: List[str], crawl_depth: int = 1,
data=data
)

def insert(self, data: Union[pd.DataFrame, Query, dict], params: dict = None):
def insert(self, data: Union[pd.DataFrame, Query, dict, list], params: dict = None):
"""
Insert data to knowledge base

Expand All @@ -176,9 +189,18 @@ def insert(self, data: Union[pd.DataFrame, Query, dict], params: dict = None):
if isinstance(data, dict):
data = [data]
elif isinstance(data, pd.DataFrame):
data = data.to_dict('records')
else:
raise ValueError("Unknown data type, accepted types: DataFrame, Query, dict")
for df in split_data(data, MAX_INSERT_SIZE):
data = df.to_dict('records')
self.insert(data, params=params)
return
elif not isinstance(data, list):
raise ValueError("Unknown data type, accepted types: DataFrame, Query, dict, list")

# chunking a big input data
if len(data) > MAX_INSERT_SIZE:
for chunk in split_data(data, MAX_INSERT_SIZE):
self.insert(chunk, params=params)
return

data = {'rows': data}
if params:
Expand Down