deeprvat.metrics

Module Contents

Classes

RSquared

Calculates the R-squared (coefficient of determination) between predictions and targets.

PearsonCorr

Calculates the Pearson correlation coefficient between burdens and targets.

PearsonCorrTorch

Calculates the Pearson correlation coefficient between burdens and targets using PyTorch tensor operations.

AveragePrecisionWithLogits

Calculates the average precision score between logits and targets.

Data

logger

API

deeprvat.metrics.logger = 'getLogger(...)'
class deeprvat.metrics.RSquared

Calculates the R-squared (coefficient of determination) between predictions and targets.

Initialization

__call__(preds: torch.tensor, targets: torch.tensor)

Calculate R-squared value between two tensors.

Parameters:

  • preds (torch.tensor) – Tensor containing predicted values.

  • targets (torch.tensor) – Tensor containing target values.

Returns:

R-squared value.

Return type:

torch.tensor

class deeprvat.metrics.PearsonCorr

Calculates the Pearson correlation coefficient between burdens and targets.

Initialization

__call__(burden, y)

Calculate Pearson correlation coefficient.

Parameters:

  • burden (torch.tensor) – Tensor containing burden values.

  • y (torch.tensor) – Tensor containing target values.

Returns:

Pearson correlation coefficient.

Return type:

float

class deeprvat.metrics.PearsonCorrTorch

Calculates the Pearson correlation coefficient between burdens and targets using PyTorch tensor operations.

Initialization

__call__(burden, y)

Calculate Pearson correlation coefficient using PyTorch tensor operations.

Parameters:

  • burden (torch.tensor) – Tensor containing burden values.

  • y (torch.tensor) – Tensor containing target values.

Returns:

Pearson correlation coefficient.

Return type:

torch.tensor

calculate_pearsonr(x, y)
class deeprvat.metrics.AveragePrecisionWithLogits

Calculates the average precision score between logits and targets.

Initialization

__call__(logits, y)

Calculate average precision score.

Parameters:

  • logits (torch.tensor) – Tensor containing logits.

  • y (torch.tensor) – Tensor containing target values.

Returns:

Average precision score.

Return type:

float