tmatchn performs efficient and flexible crossmatching
between multiple tables. It can match rows on the basis of their relative
position in the sky, or alternatively using many other criteria such as
separation in in some isotropic or anisotropic Cartesian space, identity of
a key value, or some combination of these; the full range of match criteria
is dicussed in SUN/256.
Since the match criteria define what counts as a match between two
objects, it is not immediately obvious what is meant by a multi-table match.
In fact the command can work in one of two distinct modes, controlled by the
multimode parameter. In pairs mode, one table (by default the
first input table) is designated the reference table, and pair matches
between each of the other tables and that one are identified. In
group mode groups of objects from all the input tables are
identified, as discussed in SUN/256. Currently, in both cases an output
matched row cannot contain more than one object from each input table.
Options for output of multiple rows per input table per match may be
forthcoming in future releases if there is demand.
tmatchn is intended for use with more than two input tables
- see tmatch1 and tmatch2 for 1- and 2-table crossmatching
respectively.
- nin=<count>
The number of input tables for this task. For each of the
input tables N there will be associated parameters
ifmtN,
inN
and
icmdN.
- ifmtN=<in-format>
Specifies the format of input table #N as specified by
parameter
inN. The known formats are listed in SUN/256. This flag can
be used if you know what format your table is in. If it has the special value
(auto) (the default), then an attempt will be made to detect the format
of the table automatically. This cannot always be done correctly however, in
which case the program will exit with an error explaining which formats were
attempted. This parameter is ignored for scheme-specified tables.
- inN=<tableN>
The location of input table #N. This may take one of the
following forms:
- A filename.
- A URL.
- The special value "-", meaning standard input. In this
case the input format must be given explicitly using the ifmtN
parameter. Note that not all formats can be streamed in this way.
- A scheme specification of the form
:<scheme-name>:<scheme-args>.
- A system command line with either a "<" character at
the start, or a "|" character at the end
("<syscmd" or "syscmd|"). This
executes the given pipeline and reads from its standard output. This will
probably only work on unix-like systems.
In any case, compressed data in one of the supported compression formats (gzip,
Unix compress or bzip2) will be decompressed transparently.
- icmdN=<cmds>
Specifies processing to be performed on input table #N as
specified by parameter
inN, before any other processing has taken
place. The value of this parameter is one or more of the filter commands
described in SUN/256. If more than one is given, they must be separated by
semicolon characters (";"). This parameter can be repeated multiple
times on the same command line to build up a list of processing steps. The
sequence of commands given in this way defines the processing pipeline which
is performed on the table.
Commands may alteratively be supplied in an external file, by
using the indirection character '@'. Thus a value of
"@filename" causes the file filename to be read for
a list of filter commands to execute. The commands in the file may be
separated by newline characters and/or semicolons, and lines which are blank
or which start with a '#' character are ignored.
- ocmd=<cmds>
Specifies processing to be performed on the output table,
after all other processing has taken place. The value of this parameter is one
or more of the filter commands described in SUN/256. If more than one is
given, they must be separated by semicolon characters (";"). This
parameter can be repeated multiple times on the same command line to build up
a list of processing steps. The sequence of commands given in this way defines
the processing pipeline which is performed on the table.
Commands may alteratively be supplied in an external file, by
using the indirection character '@'. Thus a value of
"@filename" causes the file filename to be read for
a list of filter commands to execute. The commands in the file may be
separated by newline characters and/or semicolons, and lines which are blank
or which start with a '#' character are ignored.
- omode=out|meta|stats|count|checksum|cgi|discard|topcat|samp|tosql|gui
The mode in which the result table will be output. The
default mode is
out, which means that the result will be written as a
new table to disk or elsewhere, as determined by the
out and
ofmt parameters. However, there are other possibilities, which
correspond to uses to which a table can be put other than outputting it, such
as displaying metadata, calculating statistics, or populating a table in an
SQL database. For some values of this parameter, additional parameters
(
<mode-args>) are required to determine the exact behaviour.
Possible values are
- out
- meta
- stats
- count
- checksum
- cgi
- discard
- topcat
- samp
- tosql
- gui
Use the
help=omode flag or see SUN/256 for more information.
- out=<out-table>
The location of the output table. This is usually a
filename to write to. If it is equal to the special value "-" (the
default) the output table will be written to standard output.
This parameter must only be given if omode has its default
value of "out".
- ofmt=<out-format>
Specifies the format in which the output table will be
written (one of the ones in SUN/256 - matching is case-insensitive and you can
use just the first few letters). If it has the special value
"
(auto)" (the default), then the output filename will be
examined to try to guess what sort of file is required usually by looking at
the extension. If it's not obvious from the filename what output format is
intended, an error will result.
This parameter must only be given if omode has its default
value of "out".
- multimode=pairs|group
Defines what is meant by a multi-table match. There are
two possibilities:
- pairs: Each output row corresponds to a single row of the reference
table (see parameter iref) and contains entries from other tables
which are pair matches to that. If a reference table row matches multiple
rows from one of the other tables, only the best one is included.
- group: Each output row corresponds to a group of entries from the
input tables which are mutually linked by pair matches between them. This
means that although you can get from any entry to any other entry via one
or more pair matches, there is no guarantee that any entry is a pair match
with any other entry. No table has privileged status in this case. If
there are multiple entries from a given table in the match group, an
arbitrary one is chosen for inclusion (there is no unique way to select
the best). See SUN/256 for more discussion.
Note that which rows actually appear in the output is also influenced by the
joinN parameter.
- iref=<table-index>
If
multimode=
pairs this parameter gives the
index of the table in the input table list which is to serve as the reference
table (the one which must be matched by other tables). Ignored in other modes.
Row ordering in the output table is usually tidiest if the default
setting of 1 is used (i.e. if the first input table is used as the reference
table).
- matcher=<matcher-name>
Defines the nature of the matching that will be
performed. Depending on the name supplied, this may be positional matching
using celestial or Cartesian coordinates, exact matching on the value of a
string column, or other things. A list and explanation of the available
matching algorithms is given in SUN/256. The value supplied for this parameter
determines the meanings of the values required by the
params,
values* and
tuning parameter(s).
- params=<match-params>
Determines the parameters of this match. This is
typically one or more tolerances such as error radii. It may contain zero or
more values; the values that are required depend on the match type selected by
the
matcher parameter. If it contains multiple values, they must be
separated by spaces; values which contain a space can be 'quoted' or
"quoted".
- tuning=<tuning-params>
Tuning values for the matching process, if appropriate.
It may contain zero or more values; the values that are permitted depend on
the match type selected by the
matcher parameter. If it contains
multiple values, they must be separated by spaces; values which contain a
space can be 'quoted' or "quoted". If this optional parameter is not
supplied, sensible defaults will be chosen.
- valuesN=<expr-list>
Defines the values from table N which are used to
determine whether a match has occurred. These will typically be coordinate
values such as RA and Dec and perhaps some per-row error values as well,
though exactly what values are required is determined by the kind of match as
determined by
matcher. Depending on the kind of match, the number and
type of the values required will be different. Multiple values should be
separated by whitespace; if whitespace occurs within a single value it must be
'quoted' or "quoted". Elements of the expression list are commonly
just column names, but may be algebraic expressions calculated from zero or
more columns as explained in SUN/256.
- joinN=default|match|nomatch|always
Determines which rows from input table N are included in
the output table. The matching algorithm determines which of the rows in each
of the input tables correspond to which rows in the other input tables, and
this parameter determines what to do with that information.
The default behaviour is that a row will appear in the output
table if it represents a match of rows from two or more of the input tables.
This can be altered on a per-input-table basis however by choosing one of
the non-default options below:
- match: Rows are included only if they contain an entry from input
table N.
- nomatch: Rows are included only if they do not contain an entry
from input table N.
- always: Rows are included if they contain an entry from input table
N (overrides any match and nomatch settings of other tables).
- default: Input table N has no special effect on whether rows are
included.
- fixcols=none|dups|all
Determines how input columns are renamed before use in
the output table. The choices are:
- none: columns are not renamed
- dups: columns which would otherwise have duplicate names in the
output will be renamed to indicate which table they came from
- all: all columns will be renamed to indicate which table they came
from
If columns are renamed, the new ones are determined by
suffix*
parameters.
- suffixN=<label>
If the
fixcols parameter is set so that input
columns are renamed for insertion into the output table, this parameter
determines how the renaming is done. It gives a suffix which is appended to
all renamed columns from table N.
- progress=none|log|time|profile
Determines whether information on progress of the match
should be output to the standard error stream as it progresses. For lengthy
matches this is a useful reassurance and can give guidance about how much
longer it will take. It can also be useful as a performance diagnostic.
The options are:
- none: no progress is shown
- log: progress information is shown
- time: progress information and some time profiling information is
shown
- profile: progress information and limited time/memory profiling
information are shown
- runner=parallel|parallel<n>|parallel-all|sequential|classic|partest
Selects the threading implementation. The options are
currently:
- parallel: uses multithreaded implementation for large tables, with
default parallelism, which is the smaller of 6 and the number of available
processors
- parallel<n>: uses multithreaded implementation for large
tables, with parallelism given by the supplied value <n>
- parallel-all: uses multithreaded implementation for large tables,
with a parallelism given by the number of available processors
- sequential: uses multithreaded implementation but with only a
single thread
- classic: uses legacy sequential implementation
- partest: uses multithreaded implementation even when tables are
small
The
parallel* options should normally run faster than
sequential
or
classic (which are provided mainly for testing purposes), at least
for large matches and where multiple processing cores are available.
The default value "parallel" is currently limited
to a parallelism of 6 since larger values yield diminishing returns given
that some parts of the matching algorithms run sequentially (Amdahl's Law),
and using too many threads can sometimes end up doing more work or impacting
on other operations on the same machine. But you can experiment with other
concurrencies, e.g. "parallel16" to run on 16 cores (if
available) or "parallel-all" to run on all available
cores.
The value of this parameter should make no difference to the
matching results. If you notice any discrepancies please report
them.