Struct FingerprintHistogram 

pub struct FingerprintHistogram { /* private fields */ }

given a collection of busfiles, FingerPrintHistogram keeps track of the number of times we see a molecule (CB/UMI/gene) in a particular distribution across the busfiles. Instead recording individual CUGs, we aggregate all the ones having the same fingerprint across samples, i.e. we create a histogram of fingerprints

Implementations

impl FingerprintHistogram

pub fn get_samplenames(&self) -> Vec<String>

pub fn get_histogram(&self) -> HashMap<Vec<u32>, usize>

pub fn from_csv(filename: &str) -> Self

load histogram from disk

pub fn new(sample_order: &[String]) -> Self

pub fn add( &mut self, record_dict: HashMap<String, Vec<BusRecord>>, ecmapper_dict: &HashMap<String, &Ec2GeneMapper>, )

adds a CUG (observed over several busfiles) into the histogram

pub fn to_dataframe(&self) -> DataFrame

turn the fingerprint histgram into a polars Dataframe

pub fn to_csv(&self, outfile: &str)

save the histogram to disk

pub fn to_csv2(&self, outfile: &str)

save the histogram to disk

pub fn estimate_sihr(&self) -> f64

Estimate SIHR from the histogram via binomial regression TODO: the paper suggests to filter some values:

• remove samples (rows over r) with less than 10 observations (m_r <10)
• remove samples with r>25

Trait Implementations

impl Debug for FingerprintHistogram

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.