NAME

stats-select - Select columns and rows from a statistics table.

SYNOPSIS

stats-select [OPTIONS] [column ...] < data-file

DESCRIPTION

Read a tab-separated statistics table or MCON log from standard input and write a tab-separated table to standard output. Positional column names choose the columns to keep, in their original order. With --remove, they instead identify columns to omit. With no column names, all columns are retained.

Column arguments may also be one-based inclusive numeric ranges, such as 2:4, :3, or 5:. Row conditions refer to the input table, so a column used for filtering need not appear in the output. Each --select condition requires an exact key=value match; repeated conditions must all match.

OPTIONS

-h, --help

Print usage information and exit.

--no-header

Omit the output line of column names.

-s key=value, --select key=value

Keep only rows with the given value in the named input column. May be repeated.

-r, --remove

Remove the listed columns instead of keeping them.

EXAMPLES

Keep the likelihood column for rows where chain=1:

stats-select --select chain=1 likelihood < samples.tsv > likelihood.tsv

Remove the second and third columns:

stats-select --remove 2:3 < samples.tsv > remaining.tsv

EXIT STATUS

Returns 0 on success and 1 on an error.

SEE ALSO

stats-cat(1), stats-merge(1)

REPORTING BUGS

BAli-Phy online help: http://www.bali-phy.org/docs.php.

Please send bug reports to bali-phy-users@googlegroups.com.