dsf.utils.validators module

Validation scripts.

dsf.utils.validators.as_1d_float_array(values, name, *, min_size=1)

Return values as a finite one-dimensional float array.

Parameters:

• values (Any) – Input values.

• name (str) – Name used in error messages.

• min_size (int) – Minimum allowed number of values.

Returns:

Finite one-dimensional float array.

Raises:

ValueError – If the input is not one-dimensional, too short, or non-finite.

Return type:

ndarray[tuple[Any, …], dtype[float64]]

dsf.utils.validators.as_2d_float_array(values, name)

Return values as a finite two-dimensional float array.

Parameters:

• values (Any) – Input values.

• name (str) – Name used in error messages.

Returns:

Finite two-dimensional float array.

Raises:

ValueError – If the input is not two-dimensional or contains non-finite values.

Return type:

ndarray[tuple[Any, …], dtype[float64]]

dsf.utils.validators.is_non_negative_integer(value)

Return whether a value is a non-negative integer.

Parameters:

value (float | int)

Return type:

bool

dsf.utils.validators.normalize_axis(axis, ndim)

Return a non-negative axis index for an array with ndim dimensions.

Parameters:

• axis (int) – Axis index, allowing negative indexing.

• ndim (int) – Number of dimensions of the target array.

Returns:

Equivalent non-negative axis index.

Raises:

ValueError – If ndim is not positive or if axis is out of bounds.

Return type:

int

dsf.utils.validators.occupied_redshift_range_from_nz(z, nz, *, threshold=0.0, name='nz')

Return the redshift range occupied by a redshift distribution.

The occupied range is defined by the first and last redshift values where the redshift distribution is above threshold. This is useful when an approximate tomographic-bin extent is needed but explicit bin edges are not available.

Parameters:

• z (Any) – Redshift grid.

• nz (Any) – Redshift distribution evaluated on z.

• threshold (float) – Minimum n(z) value used to define the occupied range.

• name (str) – Name of the redshift distribution used in error messages.

Returns:

Lower and upper occupied redshift values.

Raises:

ValueError – If the redshift distribution is invalid, if threshold is negative, or if fewer than two redshift cells are occupied.

Return type:

tuple[float, float]

dsf.utils.validators.redshift_window_mask(z, *, z_min=None, z_max=None)

Return a mask selecting redshifts inside optional window limits.

Parameters:

• z (ndarray[tuple[Any, ...], dtype[float64]])

• z_min (float | None)

• z_max (float | None)

Return type:

ndarray[tuple[Any, …], dtype[bool]]

dsf.utils.validators.validate_1d_pair(x, y, *, x_name='x', y_name='y')

Validate two matching one-dimensional arrays.

Parameters:

• x (Any) – First array. This array must be strictly increasing.

• y (Any) – Second array.

• x_name (str) – Name of the first array for error messages.

• y_name (str) – Name of the second array for error messages.

Raises:

ValueError – If arrays are not finite, one-dimensional, matching, or if x is not strictly increasing.

Return type:

None

dsf.utils.validators.validate_finite_scalar(value, name)

Validate that a scalar is finite.

Parameters:

• value (float) – Scalar value to validate.

• name (str) – Name used in error messages.

Raises:

ValueError – If value is not finite.

Return type:

None

dsf.utils.validators.validate_forecast_vector_and_covariance(data_vector, covariance, *, data_vector_name='data_vector', covariance_name='covariance')

Validate and return a forecast data vector and matching covariance.

Parameters:

• data_vector (Any) – Forecast data vector.

• covariance (Any) – Forecast covariance matrix.

• data_vector_name (str) – Name of the data vector used in error messages.

• covariance_name (str) – Name of the covariance matrix used in error messages.

Returns:

Validated one-dimensional data vector and two-dimensional covariance matrix.

Raises:

ValueError – If the data vector or covariance is invalid, or if the covariance shape does not match the data-vector length.

Return type:

tuple[ndarray[tuple[Any, …], dtype[float64]], ndarray[tuple[Any, …], dtype[float64]]]

dsf.utils.validators.validate_integration_axis(values, x, *, axis=-1)

Validate that x matches an integration axis of values.

Parameters:

• values (Any) – Array of values to integrate.

• x (Any) – Coordinate array for the integration axis.

• axis (int) – Integration axis.

Raises:

ValueError – If x is invalid or does not match the chosen axis.

Return type:

None

dsf.utils.validators.validate_integration_params(params)

Validate lens-magnification integration settings.

Parameters:

params (dict[str, Any]) – Integration settings dictionary.

Raises:

ValueError – If any setting is invalid.

Return type:

None

dsf.utils.validators.validate_joint_covariance_blocks(cov_gm_gm, cov_gg_gg, cov_gm_gg)

Validate and return covariance blocks for a joint Delta Sigma covariance.

Parameters:

• cov_gm_gm (Any) – Auto-covariance block for the galaxy-matter signal.

• cov_gg_gg (Any) – Auto-covariance block for the galaxy-galaxy signal.

• cov_gm_gg (Any) – Cross-covariance block between the two signals.

Returns:

Validated covariance blocks as finite float arrays.

Raises:

ValueError – If any block has an invalid shape.

Return type:

tuple[ndarray[tuple[Any, …], dtype[float64]], ndarray[tuple[Any, …], dtype[float64]], ndarray[tuple[Any, …], dtype[float64]]]

dsf.utils.validators.validate_nonnegative_1d_array(values, name, *, min_size=1)

Return values as a non-negative one-dimensional float array.

Parameters:

• values (Any) – Input values.

• name (str) – Name used in error messages.

• min_size (int) – Minimum allowed number of values.

Returns:

Non-negative one-dimensional float array.

Raises:

ValueError – If the input is invalid or contains negative values.

Return type:

ndarray[tuple[Any, …], dtype[float64]]

dsf.utils.validators.validate_nonnegative_scalar(value, name)

Validate that a scalar is finite and non-negative.

Parameters:

• value (float) – Scalar value to validate.

• name (str) – Name used in error messages.

Raises:

ValueError – If value is not finite or is negative.

Return type:

None

dsf.utils.validators.validate_parameter_names(parameter_names, theta0, *, parameter_names_name='parameter_names', theta0_name='theta0')

Validate parameter names against a fiducial parameter vector.

Parameters:

• parameter_names (Any) – Names associated with the forecast parameters.

• theta0 (Any) – Fiducial parameter vector.

• parameter_names_name (str) – Name of the parameter-name sequence used in error messages.

• theta0_name (str) – Name of the fiducial parameter vector used in error messages.

Returns:

Parameter names as a tuple of strings.

Raises:

• ValueError – If the number of names does not match the number of fiducial parameters.

• TypeError – If any parameter name is not a string.

Return type:

tuple[str, …]

dsf.utils.validators.validate_positive_1d_array(values, name, *, min_size=1)

Return values as a positive one-dimensional float array.

Parameters:

• values (Any) – Input values.

• name (str) – Name used in error messages.

• min_size (int) – Minimum allowed number of values.

Returns:

Positive one-dimensional float array.

Raises:

ValueError – If the input is invalid or contains non-positive values.

Return type:

ndarray[tuple[Any, …], dtype[float64]]

dsf.utils.validators.validate_positive_scalar(value, name)

Validate that a scalar is finite and positive.

Parameters:

• value (float) – Scalar value to validate.

• name (str) – Name used in error messages.

Raises:

ValueError – If value is not finite or is not positive.

Return type:

None

dsf.utils.validators.validate_positive_strictly_increasing_1d_array(values, name, *, min_size=2)

Return values as a positive strictly increasing one-dimensional array.

This is useful for physical coordinate grids such as wavenumber, radius, or frequency grids where values must be finite, positive, and ordered.

Parameters:

• values (Any) – Input coordinate values.

• name (str) – Name used in error messages.

• min_size (int) – Minimum allowed number of values.

Returns:

Positive strictly increasing one-dimensional float array.

Raises:

ValueError – If the input is invalid, non-positive, or not strictly increasing.

Return type:

ndarray[tuple[Any, …], dtype[float64]]

dsf.utils.validators.validate_power_spectrum_inputs(k, pk, *, k_name='k', pk_name='pk')

Validate a tabulated one-dimensional power spectrum.

Parameters:

• k (Any) – Wavenumber grid.

• pk (Any) – Power-spectrum values evaluated on k.

• k_name (str) – Name of the wavenumber array for error messages.

• pk_name (str) – Name of the power-spectrum array for error messages.

Raises:

ValueError – If arrays are not finite, one-dimensional, matching, positive in k, or strictly increasing in k.

Return type:

None

dsf.utils.validators.validate_redshift_distribution(z, nz, *, name='nz')

Validate and return a redshift distribution.

Parameters:

• z (Any) – Redshift grid.

• nz (Any) – Redshift distribution evaluated on z.

• name (str) – Name used in error messages.

Returns:

Redshift grid and redshift distribution as float arrays.

Raises:

ValueError – If the inputs are invalid.

Return type:

tuple[ndarray[tuple[Any, …], dtype[float64]], ndarray[tuple[Any, …], dtype[float64]]]

dsf.utils.validators.validate_redshift_distribution_support(z, nz, *, name='redshift_distribution', trim_edge_points=0)

Validate and select the positive support of a redshift distribution.

Leading and trailing zero-density points are removed. Interior non-positive values inside the selected support are rejected.

dsf.utils.validators.validate_redshift_edges(z_min, z_max, *, z_min_name='z_min', z_max_name='z_max')

Validate lower and upper redshift edges.

Parameters:

• z_min (float) – Lower redshift edge.

• z_max (float) – Upper redshift edge.

• z_min_name (str) – Name of the lower edge for error messages.

• z_max_name (str) – Name of the upper edge for error messages.

Raises:

ValueError – If the redshift edges are invalid.

Return type:

None

dsf.utils.validators.validate_redshift_pair(z_lens, z_source)

Validate a lens/source redshift pair.

Parameters:

• z_lens (float) – Lens redshift.

• z_source (float) – Source redshift.

Raises:

ValueError – If redshifts are invalid or the source is not behind the lens.

Return type:

None

dsf.utils.validators.validate_scale_factor(a)

Validate a cosmological scale factor.

Parameters:

a (float) – Scale factor.

Raises:

ValueError – If a is not in the interval 0 < a <= 1.

Return type:

None

dsf.utils.validators.validate_strictly_increasing(values, name, *, min_size=2)

Return values as a strictly increasing one-dimensional float array.

Parameters:

• values (Any) – Input values.

• name (str) – Name used in error messages.

• min_size (int) – Minimum allowed number of values.

Returns:

Strictly increasing one-dimensional float array.

Raises:

ValueError – If the input is invalid or is not strictly increasing.

Return type:

ndarray[tuple[Any, …], dtype[float64]]