Data exclusion

This section covers situations where one may want to selectively remove information from PopData.

Exclude certain things

exclude(data::PopData, kwargs...)
omit(data::PopData, kwargs...)
remove(data::PopData, kwargs...)

Returns a new PopData object without the criteria provided. Input can be a single item or, or an array of items. You will be informed you if your criteria were not found in the PopData. Using standard Julia conventions, exclude() will create a copy of your PopData excluding the specific criteria, whereas exclude!() will edit the input PopData in-place.

Keyword Arguments

Everything gets converted to string, so Symbol works too if you want to cut down on keystrokes. Integers work too if things are named "1", "2", etc.

• locus::Union{String, Vector{String}}: locus or loci you want to remove from the PopData
• population::Union{String, Vector{String}}: population(s) you want to remove from the PopData
• name::Union{String, Vector{String}}: sample(s) you want to remove from the PopData

samples

julia> fewer_sharks = exclude(sharks, name = "cc_001")
PopData{Diploid, 2209 SNP loci}
  Samples: 211
  Populations: 7
  Other info: ["longitude", "latitude"]

julia> lots_fewer_sharks = remove(sharks, name = ["cc_001", "cc_002", "cc_003"])
PopData{Diploid, 2209 SNP loci}
  Samples: 209
  Populations: 7
  Other info: ["longitude", "latitude"]

loci

julia> fewer_shark_loci = exclude(sharks, locus = "contig_475")
PopData{Diploid, 2208 SNP loci}
  Samples: 212
  Populations: 7
  Other info: ["longitude", "latitude"]

julia> lots_fewer_loci = remove(sharks, locus = ["contig_475", "contig_2784", "contig_8065"])
PopData{Diploid, 2206 SNP loci}
  Samples: 212
  Populations: 7
  Other info: ["longitude", "latitude"]

populations

julia> fewer_shark_pops = exclude(sharks, population = "Georgia")
PopData{Diploid, 2208 SNP loci}
  Samples: 182
  Populations: 6
  Other info: ["longitude", "latitude"]

julia> lots_fewer_pops = remove(sharks, population = ["FloridaKeys", "MideastGulf"])
PopData{Diploid, 2209 SNP loci}
  Samples: 119
  Populations: 5
  Other info: ["longitude", "latitude"]

combination

julia> tiny_shark = exclude(sharks, locus = "contig_475", name = ["cc_001", "neg_021", "mango_111"], population = ["CapeCanaveral", "kiwi"])
Notices:
  sample "mango_111" not found
  population "kiwi" not found

PopData{Diploid, 2208 SNP loci}
  Samples: 190
  Populations: 6
  Other info: ["longitude", "latitude"]

The in-place variant exclude!() follows all the same syntax as exclude(), therefore all examples above would be identical for exclude!().

alias functions

The exclusion commands are interchangeable with omit and remove, both with and without the bang (!). This was done so you can use the function comfortably without needing to remember the specific name to perform it. Maybe you just prefer the word omit to remove. We're not here to judge. The examples below use any combination of omit, remove, and exclude.

Keep only certain things

keep(data::PopData, kwargs...)

Returns a new PopData object keeping only the occurrences of the specified keyword. All values are converted to String for filtering, so Symbol and numbers will also work.

Keyword Arguments

• locus::Union{String, Vector{String}}: locus or loci you want to keep in the PopData
• population::Union{String, Vector{String}}: population(s) you want to keep in the PopData
• name::Union{String, Vector{String}}: sample(s) you want to keep in the PopData.

Examples

cats = @nancycats;
keep(cats, population = 1:5)
keep(cats, name = ["N100", "N102", "N211"])
keep(cats, locus = [:fca8, "fca37"])

Remove monomorphic loci

While included in the file parsers by default, you may want to do this manually with dropmonomorphic, which returns a new PopData object excluding any monomorphic loci. You can use the mutable version dropmonomorphic! to edit a PopData object in-place.

dropmonomorphic(::PopData)
dropmonomorphic!(::PopData)

Remove multiallelic markers

If your data isn't biallelic, you may want to remove multi-allelic markers for certain analyses (for example, the Hudson pairwise FST method requires biallelic data). For that we have dropmultiallelic, which returns a new PopData object, and the mutable version dropmultiallelic!, which edits a PopData object in-place.

dropmultiallelic(::PopData)
dropmultiallelic(::PopData)

Example

julia> sharks = @gulfsharks
PopData{Diploid, 2209 SNP loci}
  Samples: 212
  Populations: 7
  Other info: ["longitude", "latitude"]


julia> dropmultiallelic(sharks)
[ Info: Removing 258 multialleic loci
PopData{Diploid, 1951 SNP loci}
  Samples: 212
  Populations: 7
  Other info: ["longitude", "latitude"]