Release History¶

v0.11.0 (2021-07-21)¶

Add a Region scalar with support for loading from BED files and filtering

Fixed an issue reporting the size of GenotypeDtype in memory

Change genotype scores to be a unit8 (255=missing) instead of float64, to save ~70% of memory usage

Fix a side-effect of to_plink that swapped allele idxs which could affect edge encoding

Remove fancy new pandas dtypes from edge encoding (and rename weighted encoding to edge encoding).

Improve randomness handling in BAMS and fix calculations to ensure they work as expected (especially quantitative)

Return float values for encoding instead of newer (and less compatible) pandas types

Specify only alternate allele frequencies when generating random genotypes, to avoid float rounding problems

The genomics DataFrame Accessor no longer requires that all columns in the DataFrame are backed by a GenotypeArray

Four encoding methods added to GenotypeArray: encode_additive, encode_recessive, encode_dominant, and encode_codominant
A genotype Series accessor has been added. Thus far it includes the encoding methods.
Variants now track which allele is considered the reference allele

I/O Tests are added for Plink and VCF. Note that VCF requires htslib, so tests are skipped on Windows
Encoding tests added

Improved.

Pandas ExtensionArray tests are passing. Note: a few can’t pass until it’s possible to define dtype as scalar (See Pandas #33825)

Some initial documentation added

Initial Release