Skip to content

Latest commit

 

History

History
72 lines (61 loc) · 2.47 KB

countBy.md

File metadata and controls

72 lines (61 loc) · 2.47 KB
Raw

Recreate countBy operator for channels consisting of single elements or tuples

In DSL1, the first version of the Domain-Specific Language of Nextflow, there used to be a channel operator called countBy. It'd basically count the number of occurrences of an element in a channel, among other more interesting things. If you want to know more about it, you can see the source code before it was deprecated clicking here.

DSL doesn't have an equivalent channel operator, so if you want to do something similar, you have to use the remaining operators. Let's start with a simple example where I have a channel of letters and I want to create a new channel with these letters and the number of times they occurred. Check the source code below for a solution.

Channel
  .of('x', 'y', 'x', 'x', 'z', 'y')
  | reduce([:]) { counts, v ->
    counts[v] = (counts[v] ?: 0) + 1
    counts
  }
  | view()

If you run the snippet above, you should see something like the following:

N E X T F L O W  ~  version 23.02.0-edge
Launching `countby_with_reduce.nf` [stupefied_poisson] DSL2 - revision: fd06b3d145
[x:3, y:2, z:1]

Let's now think of something slightly trickier. What if I have a channel with tuples? Very often the elements will be different, but the first item in these elements will be an identification value in this example I'm proposing. Think of the following:

Channel
  .of(['Batman', 'DC'],
      ['Batman', 'Night knight'],
      ['Wolverine', 'Marvel'],
      ['Wolverine', 'Canadian']
  )

See what I mean? I have multiple elements that are adjectives related to a super hero. If I simply count the unique values, I will get 1 occurrence of each of the four. That's not what I want. I want the number of times we have elements whose first item is Batman, or Wolverine, and so on. See the solution below:

Channel
  .of(['Batman', 'DC'],
      ['Batman', 'Night knight'],
      ['Wolverine', 'Marvel'],
      ['Wolverine', 'Canadian']
  )
  | reduce([:]) { counts, v ->
    counts[v[0]] = (counts[v[0]] ?: 0) + 1
    counts
  }
  | view()

If you run the snippet above, you should see something like the following:

N E X T F L O W  ~  version 23.02.0-edge
Launching `countby_with_key_reduce.nf` [pedantic_joliot] DSL2 - revision: 5b2305a907
[Batman:2, Wolverine:2]