pyrolite.comp.pyrocomp (Pandas Interface)

pyrolite.comp.pyrocomp (Pandas Interface)#

class pyrolite.comp.pyrocomp(obj: DataFrame | Series)[source]#
renormalise(components: list[str] | None = None, scale: float = 100.0) → DataFrame | Series[source]#

Renormalises compositional data to ensure closure.

Parameters:

  • components (list) – Option subcompositon to renormalise to 100. Useful for the use case where compostional data and non-compositional data are stored in the same dataframe.

  • scale (float) – Closure parameter. Typically either 100 or 1.

Returns:

Renormalized dataframe.

Return type:

pandas.DataFrame

Notes

This won’t modify the dataframe in place, you’ll need to assign it to something. If you specify components, those components will be summed to 100%, and others remain unchanged.

ALR(components: list[str] | None = None, ind: int | str = -1, null_col: bool = False, label_mode: str = 'simple') → DataFrame | Series[source]#

Additive Log Ratio transformation.

Parameters:

  • ind (int | str) – Index or name of column used as denominator.

  • null_col (bool) – Whether to keep the redundant column.

Returns:

ALR-transformed array, of shape (N, D-1).

Return type:

pandas.DataFrame

inverse_ALR(ind: int | str | None = None, null_col: bool = False) → DataFrame | Series[source]#

Inverse Additive Log Ratio transformation.

Parameters:

  • ind (int | str) – Index or name of column used as denominator.

  • null_col (bool) – Whether the array contains an extra redundant column (i.e. shape is (N, D)).

Returns:

Inverse-ALR transformed array, of shape (N, D).

Return type:

pandas.DataFrame

CLR(label_mode: str = 'simple') → DataFrame | Series[source]#

Centred Log Ratio transformation.

Parameters:

label_mode (str) – Labelling mode for the output dataframe (numeric, simple, LaTeX). If you plan to use the outputs for automated visualisation and want to know which components contribute, use simple or LaTeX.

Returns:

CLR-transformed array, of shape (N, D).

Return type:

pandas.DataFrame

inverse_CLR() → DataFrame | Series[source]#

Inverse Centred Log Ratio transformation.

Returns:

Inverse-CLR transformed array, of shape (N, D).

Return type:

pandas.DataFrame

ILR(label_mode: str = 'simple') → DataFrame | Series[source]#

Isometric Log Ratio transformation.

Parameters:

label_mode (str) – Labelling mode for the output dataframe (numeric, simple, LaTeX). If you plan to use the outputs for automated visualisation and want to know which components contribute, use simple or LaTeX.

Returns:

ILR-transformed array, of shape (N, D-1)

Return type:

pandas.DataFrame

inverse_ILR(X: ndarray | None = None) → DataFrame | Series[source]#

Inverse Isometric Log Ratio transformation.

Parameters:

X (numpy.ndarray) – Optional specification for an array from which to derive the orthonormal basis, with shape (N, D).

Returns:

Inverse-ILR transformed array, of shape (N, D).

Return type:

pandas.DataFrame

boxcox(lmbda: number | None = None, lmbda_search_space: tuple[float, float] = (-1, 5), search_steps: int = 100, return_lmbda: bool = False) → DataFrame | Series[source]#

Box-Cox transformation.

Parameters:

  • lmbda (numpy.number) – Lambda value used to forward-transform values. If none, it will be calculated using the mean

  • lmbda_search_space (tuple) – Range tuple (min, max).

  • search_steps (int) – Steps for lambda search range.

Returns:

Box-Cox transformed array.

Return type:

pandas.DataFrame

inverse_boxcox(lmbda: float | None = None) → DataFrame | Series[source]#

Inverse Box-Cox transformation.

Parameters:

lmbda (float) – Lambda value used to forward-transform values.

Returns:

Inverse Box-Cox transformed array.

Return type:

pandas.DataFrame

sphere() → DataFrame | Series[source]#

Spherical coordinate transformation for compositional data.

Returns:

θ – Array of angles in radians (\((0, \pi / 2]\))

Return type:

pandas.DataFrame

inverse_sphere(variables=None) → DataFrame | Series[source]#

Inverse spherical coordinate transformation to revert back to compositional data in the simplex.

Parameters:

variables (list) – List of names for the compositional data variables, optionally specified (for when they may not be stored in the dataframe attributes through the pyrocomp() functions).

Returns:

df – Dataframe of original compositional (simplex) coordinates, normalised to 1.

Return type:

pandas.DataFrame

logratiomean(transform: Callable = <function CLR>, inverse_transform: Callable = <function inverse_CLR>) → Series[source]#

Take a mean of log-ratios along the index of a dataframe.

Parameters:

  • transform (Callable) – Log transform to use.

  • inverse_transform (Callable) – Inverse transform to use.

Returns:

Mean values as a pandas series.

Return type:

pandas.Series

Notes

Only makes sense for a dataframe.

invert_transform(**kwargs) → Series | DataFrame[source]#

Try to inverse-transform a transformed dataframe.