pin_metrics

pin_metrics(board, df_metrics, metrics_pin_name, pin_type=None, index_name='index', overwrite=False)

Update an existing pin storing model metrics over time

Parameters

Name	Type	Description	Default
`board`		Pins board	required
`df_metrics`	pd.DataFrame	Dataframe of metrics over time, such as created by `vetiver_compute_metrics()`	required
`metrics_pin_name`	str	Pin name for where the metrics are stored	required
`index_name`	str	The column in df_metrics containing the aggregated dates or datetimes. Note that this defaults to a column named “index”.	`'index'`
`overwrite`	bool	If True, overwrite any metrics for dates that exist both in the existing pin and new metrics with the new values. If False, error when the new metrics contain overlapping dates with the existing pin.	`False`

Examples

>>> import pins
>>> import vetiver
>>> df = pd.DataFrame(
... {'index': {0: pd.Timestamp('2021-01-01 00:00:00'),
...            1: pd.Timestamp('2021-01-01 00:00:00'),
...            2: pd.Timestamp('2021-01-02 00:00:00'),
...            3: pd.Timestamp('2021-01-02 00:00:00')},
...  'n': {0: 1, 1: 1, 2: 1, 3: 1},
...  'metric': {0: 'mean_squared_error',
...             1: 'mean_absolute_error',
...             2: 'mean_squared_error',
...             3: 'mean_absolute_error'},
...  'estimate': {0: 4.0, 1: 2.0, 2: 1.0, 3: 1.0}}
... )
>>> board = pins.board_temp()

>>> board.pin_write(df, "metrics", type = "csv")
>>> df = pd.DataFrame(
... {'index': {0: pd.Timestamp('2021-01-02 00:00:00'),
...            1: pd.Timestamp('2021-01-02 00:00:00'),
...            2: pd.Timestamp('2021-01-03 00:00:00'),
...            3: pd.Timestamp('2021-01-03 00:00:00')},
...  'n': {0: 1, 1: 1, 2: 1, 3: 1},
...  'metric': {0: 'mean_squared_error',
...             1: 'mean_absolute_error',
...             2: 'mean_squared_error',
...             3: 'mean_absolute_error'},
...  'estimate': {0: 4.0, 1: 6.0, 2: 2.0, 3: 1.0}}
... )
>>> vetiver.pin_metrics(
...    board=board,
...    df_metrics=df2,
...    metrics_pin_name="metrics",
...    index_name="index",
...    overwrite=True)