Skip to content

sig_score2rank_tool

Generate rank matrix from score matrix

Synopsis

sig_score2rank_tool [--ds DS] [--outfile OUTFILE] [--dim DIM] [--sort_order SORT_ORDER] [--ignore_nan IGNORE_NAN] [--as_fraction AS_FRACTION] [--as_percentile AS_PERCENTILE] [--fix_ties FIX_TIES] [--use_gctx USE_GCTX] [--read_mode READ_MODE] [--block_size BLOCK_SIZE]

Arguments

--ds DS : Input score matrix to compute ranks

--outfile OUTFILE : Output filename of rank matrix.

--dim DIM : Dimension to operate on. Default is column. Options are {column|row}

--sort_order SORT_ORDER : Sort order. Default is descend. Options are {descend|ascend}

--ignore_nan IGNORE_NAN : Ignore NaNs when ranking. Note that ignoring NaNs is slower. Default is 1

--as_fraction AS_FRACTION : Logical, returns ranks as a fraction of total rows. Default is 0

--as_percentile AS_PERCENTILE : Logical, returns ranks as a percentile of total rows. Default is 0

--fix_ties FIX_TIES : Logical, adjusts for ties (is false by default). Default is 0

--use_gctx USE_GCTX : Use GCTX file format. Default is 1

--read_mode READ_MODE : Specify how to process the input matrix. If full is specified the complete matrix is processed enmasse. If iterative the input matrix is processed in blocks. Default is full. Options are {full|iterative}

--block_size BLOCK_SIZE : size of blocks to process if the read_mode is iterative. Default is 50000

Description

Rank orders the input dataset along dimension dim in the order specified by sort_order. The rank matrix is used by connectivity analysis tools such as queryl1k to speed up the computation of enrichment based similarities.

Examples

  • Calculate ranks of input score matrix

sig_score2rank_tool --ds 'raw_data.gctx'