SEAMM Table Plug-in

A SEAMM plug-in for data tables in a flowchart.

• Free software: BSD-3-Clause

• Documentation: https://molssi-seamm.github.io/table_step/index.html

• Code: https://github.com/molssi-seamm/table_step

History

2026.10.5 – Job-level files for parallel loops

• Files named /name are written to the job’s own directory (Node.job_path) and read from it or, in an iteration of a parallel loop (loop_step 2026.10.5), from the enclosing iterations and the job. Nothing changes for other jobs.

• Requires seamm 2026.10.5.

2026.10.3 – Tables are stored in the job’s database

• Tables are now kept in the job’s database (with seamm 2026.10.3), so a job’s tables are saved with it even if they are never written to a file.

• “Append a row to” works for tables with an index column; new rows get the columns’ defaults (not empty values), and integer and boolean columns keep their types.

• “Get element of” and “Set element of” work by index value for tables whose index column holds text, and “Set element of” the current row after going past the end of the table adds the row.

• “Add columns to” uses the column name after substituting variables, and records the default.

• “Go to the next row of” past the end of a table does nothing more, so it can be used at either end of a loop.

• Values read with “Get element of” are plain Python numbers, not numpy ones.

2026.9.30 – Bugfix: appending rows to tables with text columns failed with pandas 3

• ‘Append a row to’ failed with an error naming the column (e.g. ‘SMILES’) whenever the table had a text column, because pandas 3 describes text columns differently from pandas 2. A loop over SMILES strings that appended a row each time therefore added no rows. It now works with both pandas 2 and 3, and appending to a column that the table does not have gives a clear error listing the table’s columns.

• The ‘Index column’, ‘Row’ and ‘Column’ dropdowns in the dialog offered single letters (‘c’, ‘u’, ‘r’, …) instead of ‘–none–’ or ‘current’. Fixed.

• Internal: the tests now check the rows appended in a loop, and the CI installs the package with uv from its declared requirements rather than a conda environment.

2025.6.1 – Enhancement to allow paths with directories.

• As in reading/writing structures, paths beginning with ‘/’ are relative to the root of the job, and relative paths are relative to the directory where the table step is invoked.

2023.11.10 – Bugfix: title of edit dialog was wrong

2023.10.30 – Cleaned up output

• Nothing large, just made the output properly indented, etc.

2023.7.25 – Bug fix and Enhancements

• Fixed bug with reading table using a variable for the filename, but asking for the type from the extension.

• Add ability to save tables with a frequency of other than ever call.

2023.2.15 – Bugs fixes and documentation

• Restructured documentation and moved to new theme

• Fixed bug with access rows of tables with non-integer indexes as well as “current” index

• Added support for lists of tables in pulldowns in the GUI

2021.12.22 – Improved the handling of index columns, added formats.

• Improved the handling of the index column

• Added Save as

• Added Excel and JSON formats.

2021.10.14 – Updated for Python

• Now supporting Python 3.8 and 3.9