API Reference / agent / tools / data /

read.py

Row-level dataset preview.

Functions

Func

execute_read_data

Line 9
execute_read_data(**kwargs) -> Dict[str, Any]

Read and preview actual rows from a dataset using Dataset.to_pandas().

Backs the tuiml_read_data tool.

Parameters

data
str
Dataset to read: dataset_id, file path, or built-in name (arrives via **kwargs, like all parameters below).
n_rows
int = 10
Number of rows to return, capped at 100.
mode
str = 'head'
Row selection mode: 'head', 'tail', 'sample', or 'indices'.
indices
list of int = None
Explicit row indices, used with mode='indices'.
columns
list of str = None
Subset of columns to include; unknown names are dropped.
target
str = None
Target column name, kept in the output when include_target.
include_target
bool = True
Whether to include the target column in the preview.

Returns

result
dict
On success: status ('success'), name, shape, columns, n_rows_returned, and rows (list of column -> value dicts with floats rounded to 6 places). On failure: status ('error'), error and error_type.