a3fe.analyse.analyse_set

Functionality to analyse a set of calculations and compare the result with experiment

Functions

`compute_statistic`(exp_dg, calc_dg, statistic)	Compute the desired statistic for one set of experimental and calculated values.
`compute_stats`(all_results)	Compute statistics for the passed results, generating 95 % C.I.s by bootstrapping.
`get_bootstrapped_results`(all_results[, ...])	Return n_bootstrap bootstrapped versions of the original experimental and calculated free energies.

a3fe.analyse.analyse_set.compute_statistic(exp_dg: Series, calc_dg: Series, statistic: str) → float[source]

Compute the desired statistic for one set of experimental and calculated values.

Parameters:

exp_dg (pd.Series) – The experimental free energies
calc_dg (pd.Series) – The calculated free energies
statistic (str) – The desired statistic to be calculated, from “r”, “mue”, “rmse” “rho”, or “tau”.

Returns:

The desired statistic.

Return type:

float

a3fe.analyse.analyse_set.compute_stats(all_results: DataFrame) → Dict[str, List[float]][source]

Compute statistics for the passed results, generating 95 % C.I.s by bootstrapping.

Parameters:: all_results (pd.DataFrame) – The dataframe containing all results.
Returns:: A dictionary of the computed statistics, and their upper and lower confidence bounds.
Return type:: Dict[str, List[float]]

a3fe.analyse.analyse_set.get_bootstrapped_results(all_results: DataFrame, n_bootstrap: int = 1000) → Tuple[ndarray, ndarray][source]

Return n_bootstrap bootstrapped versions of the original experimental and calculated free energies.

Parameters:

all_results (pd.DataFrame) – The dataframe containing all results.
n_bootstrap (int, optional, default = 1000) – Number of boostrap iterations to perform

Returns:

boostrapped_exp_dg (np.ndarray) – The bootstrapped experimental free energy changes
bootstrapped_calc_dg (np_ndarray) – The bootstrapped calculated free energy changes