Address order of data validation when reading from cycler

If a file is read using the `import_from_cycler` method is missing a key column like `Time [s]`, this error is only raised after the `.parquet` file is written as the data is read into the `pyprobe.procedure.Procedure` object. As `overwrite_existing=False` by default, subsequent calls to the function repeat the error even if the original cycler file has been modified.

Start at the import_from_cycler method and trace how the source data is read into pyprobe.procedure.Procedure and then written to the .parquet file. Confirm that missing key columns such as Time [s] are validated before writing; done means the invalid import raises the error without creating or reusing an incomplete output.

Written by the indexing model from the issue text.