NEWS

# 0.8.0

  * Function renaming:
    * `cbind_SNPs` --> `cbind_snp_files`
    * `rbind_SNPs` --> `rbind_snp_files`
    * `mask_SNPs` --> `mask_snp_file`

  * Added support for reading SHAPEIT haps/sample files (`read.haps`)
  * Added support for reading Oxford formatted files (`read.oxford`).
  
  * `imputation_accuracy`: Standardizing with `p` is now scaling with the
    *squareroot* of the variance of allele frequency. Before this, it was the
    variance.

# 0.7.1

  * Fortran routines updated to be able to handle integer IDs exceeding 
    2147483647. Passing larger integers back to R is however still not possible.
  * Added support for vcf files with package 'vcfR', but require the genotypes
    to be loaded with `read.vcfR`.
    Support is on `write.snps`, and `imputation_accuracy`.
  * `imputation_accuracy` has slight modified behaviour:
    It now dispatches to either files or matrices depending on first argument.
    Correct.pct is calculated as # correct / (elements - true missing).
    Tolerance is limited to 4 digits. I.e. a tolerance below 1e-4 is effectively
    ignored.

# 0.5.0

  * Preparing package for publication.
  * get_firstcolumn now works with several data types, but default use is still integer.

# 0.3.2

  * Added function `mergeChips` for merging multiple SNP chips into a single file.
  * Added function `read.snps` for easy reading SNP files.

# 0.3.1

  * Added function `rowconcatenate`, albeit without test routines.
  * Updated phasotogeno to handle numeric values instead of just integers.
  * phasotogeno_int is the integer version of phasotogeno (where the latter does not work with heterozygosity).
  * Bug fix: Now standardises by standard deviation, not variance.

# 0.3.0

  * Added NEWS file.
  * In `imputation_accuracy3`, row and columns correlations that are `Inf`, `-Inf`, or `NaN` are set to `NA`.

# 0.2.0

  * Added Heterozygosity function.