iCLIP.meta.quantile_row_norm

iCLIP.meta.quantile_row_norm(matrix, quantile=0.99, min_value='auto')

Normalise the rows of a matrix so that 1 corresponds to the given quantile.

Parameters:min_value – allows a value to be provided so that

the minimum normalisation value will never be smaller than this. If set to ‘auto’ then the smallest non-zero value across the whole matrix will be found and used

Parameters:

• matrix (pandas.DataFrame) – Matrix of data for which the rows are to be normalized
• quantile (float, optional) – Quantile to use as the normalizer, defaults to 0.99 thus removing only the most outlying values.
• min_value (float or "auto") – The minimum value the normalizer can take. This prevents rows being normalized by 0. If ‘auto’ then for each for the smallest non zero value is used.

Returns:

Same row and column indexes as matrix. Rows are now normalized

Return type:

pandas.DataFrame