-
|
As of now, I utilize Pandas query pipelines and craft prompts to gain context through field names. Are there recommended practices for building a 'Data Analyst' Agent that can summarize data in a specific manner? Ultimately, I aim to extract 10-15 key insights about the dataset and its trends. I am thinking having Data Analyst agent will be more powerful than writing everytihng in the query prompt. I am proficient in Python and favor Python query syntax over SQL. Thank you for your help. https://docs.llamaindex.ai/en/stable/examples/pipeline/query_pipeline_pandas/ |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 10 replies
-
|
To build a 'Data Analyst' Agent that can summarize data using Python and Pandas, follow these steps:
This approach leverages your proficiency in Pandas and Python, automating the extraction of valuable insights from datasets. For more details, refer to the PandasQueryEngine source code.
|
Beta Was this translation helpful? Give feedback.
-
|
@dosu Thank you. This is exactly I am looking here. BTW, can I pass a series of questions via qp.run? I like the way you port the list of questions in the first post. I like to do the same via qp.run process. The below is your example::::: |
Beta Was this translation helpful? Give feedback.
-
|
To process a series of questions via
Here's a code snippet to illustrate the concept with batch processing: # Assuming qp is your QP instance and it's already set up as shown previously
questions = [
"What is the average value of column_name?",
"Identify trends in column_name over time.",
"Detect outliers in column_name."
]
# Process questions in batch
for question in questions:
response = qp.run(question)
print(f"Question: {question}\nResponse: {response}\n")This approach allows you to leverage the full capabilities of the
|
Beta Was this translation helpful? Give feedback.
-
|
I am getting ValueError: Only one free input key is allowed. Please help. |
Beta Was this translation helpful? Give feedback.
-
|
To address the "ValueError: Only one free input key is allowed" issue you're encountering, it's important to ensure that your function call matches the expected input structure of the
By following these steps and ensuring your inputs are correctly structured for the function's expectations, you should be able to resolve the "ValueError: Only one free input key is allowed" and successfully process your batch of questions.
|
Beta Was this translation helpful? Give feedback.
To address the "ValueError: Only one free input key is allowed" issue you're encountering, it's important to ensure that your function call matches the expected input structure of the
process_questionsfunction or any similar function you're using. This error typically occurs when the function is designed to accept a single type of input but receives multiple inputs that it cannot process simultaneously. Here are steps to resolve the issue:Check Input Structure: Make sure you're only providing the necessary input that the function expects. If the function is designed to process queries, ensure you're not passing additional inputs that could be interpreted as free input keys.
Review …