XSV
Perform various operations over CSV/TSV tables.
URL: https://github.com/BurntSushi/xsv
Example
This wrapper can be used in the following way:
### Concatenation subcommand ###
rule test_xsv_cat_rows:
input:
table=["table.csv", "right.csv"],
output:
"xsv/catrows.csv",
log:
"logs/catrow.log",
params:
subcommand="cat rows",
extra="",
threads: 1
wrapper:
"v3.9.0/utils/xsv"
use rule test_xsv_cat_rows as test_xsv_cat_cols with:
output:
"xsv/catcols.csv",
log:
"logs/catcol.log",
params:
subcommand="cat columns",
### Count subcommand ###
rule test_xsv_count:
input:
table="table.csv",
output:
"xsv/count_csv.csv",
log:
"logs/count_csv.log",
params:
subcommand="count",
extra="",
threads: 1
wrapper:
"v3.9.0/utils/xsv"
use rule test_xsv_count as test_xsv_count_tsv_input with:
input:
table="table.tsv",
output:
"xsv/count_tsv.csv",
log:
"logs/count_tsv.log",
### Fix lengths subcommand ###
use rule test_xsv_count as test_xsv_fixlength with:
output:
"xsv/fixlength.csv",
log:
"logs/fixlength.log",
params:
subcommand="fixlengths",
extra="--length 20",
### Flatten subcommand ###
use rule test_xsv_count as test_xsv_flatten with:
output:
"xsv/flatten.csv",
log:
"logs/flatten.log",
params:
subcommand="flatten",
### Format subcommand ###
use rule test_xsv_count as test_xsv_fmt with:
output:
"xsv/fmt.tsv",
log:
"logs/fmt.log",
params:
subcommand="fmt",
### Frequency subcommand ###
use rule test_xsv_count as test_xsv_frequency with:
output:
"xsv/frequency.csv",
log:
"logs/frequency.log",
params:
subcommand="frequency",
### Headers subcommand ###
use rule test_xsv_count as test_xsv_headers with:
output:
"xsv/headers.csv",
log:
"logs/headers.log",
params:
subcommand="headers",
use rule test_xsv_cat_rows as test_xsv_headers_list with:
output:
"xsv/headers_all.csv",
log:
"logs/headers_all.log",
params:
subcommand="headers",
extra="--intersect",
### Index subcommand ###
use rule test_xsv_count as test_xsv_index with:
output:
"table.csv.idx",
log:
"logs/index.log",
params:
subcommand="index",
### Input subcommand ###
use rule test_xsv_count as test_xsv_input with:
output:
"xsv/input.csv",
log:
"logs/input.log",
params:
subcommand="input",
### Join subcommand ###
use rule test_xsv_cat_rows as test_xsv_join with:
output:
"xsv/join.csv",
log:
"logs/join.log",
params:
subcommand="join",
col1="gene_id",
col2="gene_id",
### Sample subcommand ###
use rule test_xsv_count as test_xsv_sample with:
output:
"xsv/sample.csv",
log:
"logs/sample.log",
params:
subcommand="sample",
extra="1",
### Search subcommand ###
use rule test_xsv_count as test_xsv_search with:
output:
"xsv/search.csv",
log:
"logs/search.log",
params:
subcommand="search",
extra="--select gene_id ENSG[0-9]+",
### Select subcommand ###
use rule test_xsv_count as test_xsv_select with:
output:
"xsv/select.csv",
log:
"logs/select.log",
params:
subcommand="select",
extra="3-",
### Slice subcommand ###
use rule test_xsv_count as test_xsv_slice with:
output:
"xsv/slice.csv",
log:
"logs/slice.log",
params:
subcommand="slice",
extra="-i 2",
### Sort subcommand ###
use rule test_xsv_count as test_xsv_sort with:
output:
"xsv/sort.csv",
log:
"logs/sort.log",
params:
subcommand="sort",
### Split subcommand ###
use rule test_xsv_count as test_xsv_split with:
output:
directory("xsv/split"),
log:
"logs/split.log",
params:
subcommand="split",
extra="-s 2",
use rule test_xsv_count as test_xsv_split_list with:
output:
expand("xsv/split/{nb}.csv", nb=["0", "1"]),
log:
"logs/split.log",
params:
subcommand="split",
extra="-s 1",
### Stat subcommand ###
use rule test_xsv_count as test_xsv_stats with:
output:
"xsv/stats.txt",
log:
"logs/stats.log",
params:
subcommand="stats",
### Table subcommand ###
rule test_xsv_table:
input:
table="right.csv",
output:
"xsv/table.txt",
log:
"logs/table.log",
params:
subcommand="table",
extra="",
threads: 1
wrapper:
"v3.9.0/utils/xsv"
Note that input, output and log file paths can be chosen freely.
When running with
snakemake --use-conda
the software dependencies will be automatically deployed into an isolated environment before execution.
Notes
Adding table(s) index(es) to the input file list makes many subcommands faster.
Software dependencies
xsv=0.13.0
Input/Output
Input:
table: Path to CSV/TSV table.
Output:
Path the result file / directory
Params
extra: Optional arguments for xsv. For TSV files, –delimiter is automatically set to a tabulation.subcommand: xsv subcommand among cat, count, fixlengths, flatten, fmt, frequency, headers, index, input, join, sample, search, select, slice, sort, split, stats, or table
Code
__author__ = "Thibault Dayris"
__copyright__ = "Copyright 2023, Thibault Dayris"
__email__ = "thibault.dayris@gustaveroussy.fr"
__license__ = "MIT"
import os
from snakemake.shell import shell
log = snakemake.log_fmt_shell(stdout=False, stderr=True)
subcommand = snakemake.params["subcommand"]
extra = snakemake.params.get("extra", "")
# TSV delimiter
if len(snakemake.input) == 1:
if str(snakemake.input).endswith(".tsv"):
extra += " --delimiter $'\t'"
elif all(table.endswith(".tsv") for table in snakemake.input):
extra += " --delimiter $'\t'"
# Automatic multithreading when possible
if subcommand in ["frequency", "split", "stats"]:
extra += f" --jobs {snakemake.threads}"
elif snakemake.threads > 1:
raise Warning("Only one thread is required")
# Command line building
if subcommand == "join":
shell(
"xsv {subcommand} {extra} "
"{snakemake.params.col1} {snakemake.input[0]} "
"{snakemake.params.col2} {snakemake.input[1]} "
"> {snakemake.output} {log}"
)
elif subcommand == "index":
log = snakemake.log_fmt_shell(stdout=True, stderr=True)
shell("xsv {subcommand} {extra} {snakemake.input} {log}")
elif subcommand == "split":
log = snakemake.log_fmt_shell(stdout=True, stderr=True)
outdir = snakemake.output
if len(outdir) > 1:
outdir = os.path.dirname(outdir[0])
shell("xsv {subcommand} {extra} {outdir} {snakemake.input} {log}")
else:
shell("xsv {subcommand} {extra} {snakemake.input} > {snakemake.output} {log}")