Fort99 Analyzer Analysis

fort.99 (training-set and EOS) analysis utilities.

This module provides helpers for analyzing ReaxFF training-set output stored in fort.99 files, including error inspection, two-body ENERGY term parsing, and equation-of-state (EOS) analysis for bulk modulus extraction.

Typical use cases include:

inspecting QM–FF energy differences for training targets
parsing pairwise ENERGY terms from fort.99 titles
constructing energy–volume datasets using fort.74
computing bulk modulus via a Vinet equation-of-state fit

`fort99_energy_vs_volume(fort99_handler, fort74_handler)`

Construct an energy–volume table from fort.99 and fort.74 data.

Pairwise ENERGY terms from fort.99 are matched with corresponding volumes extracted from fort.74 based on shared identifiers.

Works on

Fort99Handler + Fort74Handler — fort.99 + fort.74

Parameters:

Name	Type	Description	Default
`fort99_handler`	`TemplateHandler`	Parsed `fort.99` handler.	required
`fort74_handler`	`TemplateHandler`	Parsed `fort.74` handler providing volumes.	required

Returns:

Type	Description
`DataFrame`	Table with columns: `iden1`, `iden2`, `ffield_value`, `qm_value`, `V_iden2`.

Examples:

>>> df = fort99_energy_vs_volume(f99, f74)

`get_fort99_bulk_modulus(fort99_handler, fort74_handler, *, iden, source='ffield', shift_min_to_zero=True, flip_sign=False, dropna=True)`

Compute the bulk modulus using a Vinet EOS fit to E(V) data.

Energy–volume points are obtained from pairwise ENERGY terms in fort.99 and corresponding volumes from fort.74.

Works on

Fort99Handler + Fort74Handler — fort.99 + fort.74

Parameters:

Name	Type	Description	Default
`fort99_handler`	`TemplateHandler`	Parsed `fort.99` handler.	required
`fort74_handler`	`TemplateHandler`	Parsed `fort.74` handler.	required
`iden`	`str`	Identifier (`iden1`) for which the EOS fit is performed.	required
`source`	`('ffield', 'qm')`	Energy source used for fitting.	`"ffield"`
`shift_min_to_zero`	`bool`	Shift minimum energy to zero to improve numerical conditioning.	`True`
`flip_sign`	`bool`	Flip the sign of energies before fitting.	`False`
`dropna`	`bool`	Drop rows with NaN energy or volume values.	`True`

Returns:

Type	Description
`dict`	Dictionary containing: `iden`, `source`, `n_points`, `V0_A3`, `K0_eV_A3`, `K0_GPa`, `E0_eV`, `C`, `success`.

Examples:

>>> res = get_fort99_bulk_modulus(
...     f99, f74, iden="bulk_0", source="ffield"
... )
>>> res["K0_GPa"]

`get_fort99_data(handler, *, sortby='lineno', ascending=True)`

Retrieve fort.99 data and compute QM–FF energy differences.

A new column qm_ff_difference is added as: qm_value - ffield_value.

Works on

Fort99Handler — fort.99

Parameters:

Name	Type	Description	Default
`handler`	`TemplateHandler`	Parsed `fort.99` handler.	required
`sortby`	`str`	Column name to sort by (e.g. `error`, `lineno`).	`"lineno"`
`ascending`	`bool`	Sort order.	`True`

Returns:

Type	Description
`DataFrame`	fort.99 table including the additional `qm_ff_difference` column.

Examples:

>>> df = get_fort99_data(h, sortby="error", ascending=False)

`parse_fort99_two_body_energy_terms(handler)`

Parse pairwise ENERGY terms from the ENERGY section of fort.99.

Only ENERGY titles containing exactly two / separators are kept, corresponding to two-body interaction terms.

Works on

Fort99Handler — fort.99

Parameters:

Name	Type	Description	Default
`handler`	`TemplateHandler`	Parsed `fort.99` handler.	required

Returns:

Type	Description
`DataFrame`	ENERGY rows augmented with parsed columns: `opt1`, `iden1`, `n1`, `opt2`, `iden2`, `n2`.

Examples:

>>> df = parse_fort99_two_body_energy_terms(h)