dataguy.DataGuy

class DataGuy(max_code_history=100, model_overrides: Dict[str, Any] | None = None, auto_model_switch=True)

An automatic tool for describing, analyzing and plotting data. It uses an LLM model to generate lambda functions for various data analysis methods. Its execution checks if the code has an error, and generates it again, by feeding the old code back together with the error.

__init__(max_code_history=100, model_overrides: Dict[str, Any] | None = None, auto_model_switch=True)

Methods

__init__([max_code_history, ...])
analyze_data()	Analyzes the data stored in DataGuy.
describe_data()	Creates a description of the data, from the summary.
describe_plot(img_bytes)	Creates a detailed description from the given plot.
plot_data(column_x, column_y)	Creates a scatterplot.
recreate_plot(plot_description)	Takes in a description made by describe_plot, and tries to recreate it using the data.
set_data(obj)	Imports the data into the structure.
summarize_data()	Generates a summary of the data.
wrangle_data()	Calls the _generate_code function with the task to write a wrangler lambda function,