mirror of
https://github.com/johnkerl/miller.git
synced 2026-07-22 15:37:59 +00:00
todo
This commit is contained in:
parent
4f6572f4b2
commit
e7740e6766
1 changed files with 68 additions and 1 deletions
69
c/todo.txt
69
c/todo.txt
|
|
@ -8,7 +8,7 @@ TOP-OF-LIST SUMMARY
|
|||
|
||||
* rh/fedora/centos mlr-3.4.0
|
||||
|
||||
* explode & any other komosas/wolens from feature-request task.
|
||||
* explode/unnest/etc
|
||||
* tri-split null
|
||||
|
||||
* parameterized emit: needs spec & impl. prereq: tri-split null.
|
||||
|
|
@ -30,6 +30,73 @@ TOP-OF-LIST SUMMARY
|
|||
================================================================
|
||||
TOP-OF-LIST DETAILS
|
||||
|
||||
----------------------------------------------------------------
|
||||
* explode/unnest/etc
|
||||
|
||||
TSV:
|
||||
|
||||
x y
|
||||
1 a
|
||||
2 d,e,f
|
||||
3 g,h
|
||||
|
||||
mlr unnest --fs comma -f y would produce
|
||||
|
||||
x y
|
||||
1 a
|
||||
2 d
|
||||
2 e
|
||||
2 f
|
||||
3 g
|
||||
3 h
|
||||
|
||||
TSV:
|
||||
|
||||
x y
|
||||
1 a
|
||||
2 s=d,t=e,u=f
|
||||
3 u=g,v=h
|
||||
|
||||
maps to
|
||||
|
||||
x=1,y=a
|
||||
x=2,s=d,t=e,u=f
|
||||
x=3,u=g,v=h
|
||||
|
||||
likewise
|
||||
|
||||
timestamp here some prefixes follow, StatsType=datareporter,field1=2,field3=4
|
||||
|
||||
field1=2,field3=4
|
||||
|
||||
Being able to deal with replicates in both directions would be really useful.
|
||||
Here's a real world example I was working on today, which contains two different measures of association between pairs of
|
||||
genetic markers (rs*):
|
||||
|
||||
rs10 rs2237570,0.3,-0.72;rs117465896,0.75,-0.88
|
||||
rs10041592 rs1628466,0.21,0.92;rs1729036,0.24,1;rs79177551,0.2,0.85
|
||||
rs10033329 rs4835294,0.2,0.46
|
||||
|
||||
There are two fields, tab-delimited. Records in the second field provide the association results for all markers
|
||||
relevant to the marker in the first field. Converting this format into a "long" format requires two operators. First,
|
||||
the unnesting step I described above, which spreads the semicolon-separated replicates across new records:
|
||||
|
||||
rs10 rs2237570,0.3,-0.72
|
||||
rs10 rs117465896,0.75,-0.88
|
||||
rs10041592 rs1628466,0.21,0.92
|
||||
rs10041592 rs1729036,0.24,1
|
||||
rs10041592 rs79177551,0.2,0.85
|
||||
rs10033329 rs4835294,0.2,0.46
|
||||
|
||||
followed by a separating step (as described by @komosa) to spread the comma-separated replicates across new fields:
|
||||
|
||||
rs10 rs2237570 0.3 -0.72
|
||||
rs10 rs117465896 0.75 -0.88
|
||||
rs10041592 rs1628466 0.21 0.92
|
||||
rs10041592 rs1729036 0.24 1
|
||||
rs10041592 rs79177551 0.2 0.85
|
||||
rs10033329 rs4835294 0.2 0.46
|
||||
|
||||
----------------------------------------------------------------
|
||||
* tri-split null
|
||||
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue