The community member posted about the lack of support in llama-index for creating a query pipeline/engine over multiple Pandas dataframes, and asked if this is in the roadmap. In the comments, another community member responded that this is not really on the roadmap, but they are open to contributions. They also mentioned that Pandas in general is a security concern due to the risk of arbitrary code execution, and suggested converting to SQLite and using "text to SQL" as an alternative approach. The original poster acknowledged this suggestion and expressed interest in exploring the capabilities of "text to SQL" further, as the data in their Pandas dataframes originally came from different tables in an Azure SQL database.
apparently llama-index do no have support for creating query pipeline/engine over multiple pandas dataframe yet! Just wondering if this is in the roadmap?
hmm okk, text to sql is something new to think about. Because originally, the data of dataframes comes from different tables sitting in azure sql db. Have not yet explored the capabilities of text to sql. But thanks!
