epiout.sflognorm

Module Contents

Classes

SFLogNorm

Size factor log normalization for samples based on the geometric mean.

class epiout.sflognorm.SFLogNorm(mean_norm=True)

Size factor log normalization for samples based on the geometric mean.

Parameters

mean_norm (bool) – Whether to mean normalize the data.

fit(self, X)

Fit the size factor log normalization.

static size_factor(X)

Size factor for each sample based on the geometric mean.

Parameters

X (tf.Tensor) – Input data as counts.

static sflog(X, sf)

Size factor log transformation.

Parameters
  • X (tf.Tensor) – Input data as counts.

  • sf (tf.Tensor) – Size factor for each sample.

transform(self, X)

Transform the data with the size factor log normalization.

Parameters

X (tf.Tensor) – Input data as counts.

fit_transform(self, X)

Fit size factor using count data and transform the data.

Parameters

X (tf.Tensor) – Input data as counts.

inverse_transform(self, X)
Inverse transform the data back to counts from the size factor

log normalization values.

Parameters

X (tf.Tensor) – Input data as size factor log normalization values.