hic3defdr.util.scaling module

hic3defdr.util.scaling.conditional(data, dist, fn, n_bins=None)[source]

Applies a size factor computing function fn to data conditioning on dist, optionally binning dist into n_bins equal-number bins.

If n_bins is not None, the final size factors will be interpolated for interaction distances falling between the bins to avoid bin edge effects.

Parameters:
  • data (np.ndarray) – Rows correspond to pixels, columns correspond to replicates.
  • dist (np.ndarray) – The distance of each pixel in data.
  • fn (function) – A function that computes a size factor given some data.
  • n_bins (int, optional) – Pass an int to bin distance into this many equal-number bins.
hic3defdr.util.scaling.conditional_mor(data, dist, n_bins=None)[source]

Computes size factors for a dataset using median of ratios normalization, conditioning on distance.

Parameters:
  • data (np.ndarray) – Rows correspond to pixels, columns correspond to replicates.
  • dist (np.ndarray) – The distance of each pixel in data
  • n_bins (int, optional) – Pass an int to bin distance into this many equal-number bins.
Returns:

Matrix of size factors, per pixel (rows) and per replicate (columns).
Return type:

np.ndarray
hic3defdr.util.scaling.conditional_scaling(data, dist, n_bins=None)[source]

Computes size factors for a dataset using simple scaling normalization, conditioning on distance.

Parameters:
  • data (np.ndarray) – Rows correspond to pixels, columns correspond to replicates.
  • dist (np.ndarray) – The distance of each pixel in data.
  • n_bins (int, optional) – Pass an int to bin distance into this many equal-number bins.
Returns:

Matrix of size factors, per pixel (rows) and per replicate (columns).
Return type:

np.ndarray
hic3defdr.util.scaling.median_of_ratios(data, filter_zeros=True)[source]

Computes size factors for a dataset using the median of ratios method.

Parameters:
  • data (np.ndarray) – Rows correspond to pixels, columns correspond to replicates.
  • filter_zeros (bool) – Pass True to filter out pixels with a zero value in any replicate. Pass False to include all data.
Returns:

Vector of size factors, per replicate.
Return type:

np.ndarray
hic3defdr.util.scaling.no_scaling(data)[source]

Computes dummy size factors of 1 for each replicate.

Parameters:data (np.ndarray) – Rows correspond to pixels, columns correspond to replicates.
Returns:Vector of size factors, per replicate.
Return type:np.ndarray
hic3defdr.util.scaling.simple_scaling(data)[source]

Computes size factors for a dataset using a simple scaling method.

Parameters:data (np.ndarray) – Rows correspond to pixels, columns correspond to replicates.
Returns:Vector of size factors, per replicate.
Return type:np.ndarray