safe_write_tac
===============================================

.. py:function:: petpal.utils.time_activity_curve.safe_write_tac(filename: str, tac_data: numpy.ndarray, col_names: list[str] = None)

   Writes the data in a time-activity curve (TAC) to a file. Assumes the TAC data consists of a
   contiguous numpy array with two or three columns: time, activity, and (optionally) uncertainty.

   :param filename: Path to the file the data will be saved as.
   :type filename: str
   :param tac_data: Numpy array containing the data for the TAC. Assumes the TAC data
                    consists of a contiguous numpy array with two or three columns: time, activity, and
                    (optionally) uncertainty.
   :type tac_data: np.ndarray
   :param col_names: List of column names assigned to the time, activity, and uncertainty
                     columns in the TAC data, respectively. Must match number of columns in `tac_data`.
   :type col_names: list[str]

   :raises ValueError: If the number of columns in tac_data is not two or three, or the number of
       columns in tac_data does not match the number of columns in col_names.

   .. rubric:: Example

   .. code-block:: python

       import numpy as np
       from petpal.utils.time_activity_curve import safe_load_tac, safe_write_tac

       my_tac = safe_load_tac(filename='/path/to/tac.tsv')
       my_tac_modified = np.zeros_like(my_tac)
       my_tac_modified[0] = my_tac[0] / 60 # convert time units to hours
       my_tac_modified[1] = my_tac[1] / 37000 # convert activity units to mCi
       my_tac_modified[2] = my_tac[2] / 37000 # convert uncertainties like activity
       safe_write_tac(filename='/path/to/new_tac.tsv',
                      tac_data=my_tac_modified)