Table

​

class wandb.Table

columns: 'list[ColumnKey] | None' = None,
data: 'list[InputRow] | np.ndarray | pd.DataFrame | None' = None,
rows: 'list[InputRow] | None' = None,
dataframe: 'pd.DataFrame | None' = None,
dtype: 'Any' = None,
optional: 'bool | list[bool]' = True,
allow_mixed_types: 'bool' = False,
log_mode: 'LogMode | None' = 'IMMUTABLE'

​

Args

• columns: Names of the columns in the table. Defaults to [“Input”, “Output”, “Expected”].
• data: 2D row-oriented array of values, NumPy array, or pandas DataFrame.
• rows: 2D row-oriented array of values.
• dataframe: pandas DataFrame object used to create the table. When set, data and columns arguments are ignored.
• dtype: The expected type for the column values, used to validate the data. If not set, types are inferred from the data. It can be:
  • a single type
    • a Python built-in type such as int, str, bool, list, dict, or datetime.
    • a W&B Media type like wandb.Image declared under wandb.data_types
    • a const value
  • a list of any of the above to assign a different type to each column (should be the same length as columns)
• optional: Determines if None values are allowed. Defaults to True.
  • If a singular bool value, then the optionality is enforced for all columns specified at construction time
  • If a list of bool values, then the optionality is applied to each column - should be the same length as columns applies to all columns. A list of bool values applies to each respective column.
• allow_mixed_types: Determines if columns are allowed to have mixed types (disables type validation). Defaults to False
• log_mode: Controls how the Table is logged when mutations occur. Options:
  • “IMMUTABLE” (default): Table can only be logged once; subsequent logging attempts after the table has been mutated will be no-ops.
  • “MUTABLE”: Table can be re-logged after mutations, creating a new artifact version each time it’s logged.
  • “INCREMENTAL”: Table data is logged incrementally, with each log creating a new artifact entry containing the new data since the last log.

​

Methods

​

method Table.add_column()

self,
name: 'str',
data: 'list[Any] | np.ndarray',
optional: 'bool' = False

Arguments

• name: The unique name of the column.
• data: A column of homogeneous data.
• optional: If null-like values are permitted.

​

method Table.add_computed_columns()

self,
fn: 'Callable[[int, dict[ColumnKey, Any]], dict[str, Any]]'

Arguments

• fn: A function which accepts an index and row dict, and returns a dict representing new columns for that row, keyed by the new column names.

Examples

• ndx is an integer representing the index of the row.
• row is a dictionary keyed by existing columns.

import wandb

table = wandb.Table(columns=["x", "y"], data=[[3, 1], [4, 6]])
table.add_computed_columns(lambda ndx, row: {"diff": row["x"] - row["y"]})

​

method Table.add_data()

self, *data: 'Any'

Arguments

• data:

​

method Table.add_row()

self, *row: 'Any'

Arguments

• row:

​

method Table.cast()

self,
col_name: 'ColumnKey',
dtype: 'Any',
optional: 'bool' = False

Arguments

• col_name: The name of the column to cast.
• dtype: The target dtype.
• optional: If the column should allow Nones.

​

method Table.get_column()

self,
name: 'ColumnKey',
convert_to: "Literal['numpy'] | None" = None

Arguments

• name: The name of the column.
• convert_to: “numpy” will convert the underlying data to a NumPy object.

​

method Table.get_dataframe()

self

​

method Table.get_index()

(self)