o Nrf+@sddlZddlZddlZddlZddlZddlm Z m Z m Z m Z m Z ddlmZmZddlmZmZmZeejddd ejd ejd d d d dejddd ejdddeddgd dejdddejdddejdded dejdd d d!d"ejd#d d d$d"ejd%ed&d d'd(ejd)d*d+d d d,ejd-ed.d/gd.d0d d1ejd2d3d4ed5d6ejd7d8ed9ejd:d;ejd d d d d<d9ejd=d>ed?d d@ejdAdBd d d,ejdCdDdejdEdFd d dGd"dHdIZdS)JN) BIN_DTYPE COUNT_DTYPEcreate_from_unorderedsanitize_pixelssanitize_records)cli get_logger) parse_binsparse_field_paramparse_kv_list_param bins_pathZ BINS_PATH)metavar pixels_pathT)exists allow_dashZ PIXELS_PATH)typer cool_pathZ COOL_PATHz--formatz-fz'coo' refers to a tab-delimited sparse triplet file (bin1, bin2, count). 'bg2' refers to a 2D bedGraph-like file (chrom1, start1, end1, chrom2, start2, end2, count).coobg2)helprrequiredz --metadataz+Path to JSON file containing user metadata.)rz --assemblyz)Name of genome assembly (e.g. hg19, mm10)z--fieldaAdd supplemental value fields or override default field numbers for the specified format. Specify quantitative input fields to aggregate into value columns using the syntax ``--field =``. Optionally, append ``:`` followed by ``dtype=`` to specify the data type (e.g. float). Field numbers are 1-based. Repeat the ``--field`` option for each additional field.)rrZmultiplez--count-as-floatFz{Store the 'count' column as floating point values instead of as integers. Can also be specified using the `--field` option.)is_flagdefaultrz --one-basedzWPass this flag if the bin IDs listed in a COO file are one-based instead of zero-based.z--comment-char#z1Comment character that indicates lines to ignore.)rr show_defaultrz--no-symmetric-upperz-NztCreate a complete square matrix without implicit symmetry. This allows for distinct upper- and lower-triangle values)rrrz--input-copy-statusuniqueduplexaCopy status of input data when using symmetric-upper storage. | `unique`: Incoming data comes from a unique half of a symmetric matrix, regardless of how element coordinates are ordered. Execution will be aborted if duplicates are detected. `duplex`: Incoming data contains upper- and lower-triangle duplicates. All lower-triangle input elements will be discarded! | If you wish to treat lower- and upper-triangle input data as distinct, use the ``--no-symmetric-upper`` option instead. )rrrrz --chunksizez-czSize in number of lines/records of data chunks to read and process from the input stream at a time. These chunks will be saved as temporary partial coolers and then merged.i-1)rrrz --mergebufzgTotal number of records to buffer per epoch of merging data. Defaults to the same value as `chunksize`.)rrz --temp-dirzaCreate temporary files in a specified directory. Pass ``-`` to use the platform default temp dir.)rZ file_okayZdir_okayrz --max-mergez3Maximum number of chunks to merge in a single pass.)rrrrz--no-delete-tempz,Do not delete temporary files when finished.z--storage-optionszOptions to modify the data filter pipeline. Provide as a comma-separated list of key-value pairs of the form 'k1=v1,k2=v2,...'. See http://docs.h5py.org/en/stable/high/dataset.html#filter-pipeline for more details.z--appendz-az_Pass this flag to append the output cooler to an existing file instead of overwriting the file.c( s\tt}t|\}}| dur| } | }d}|r$| dkrd}n| dkr$d}|durAt| }t|}Wdn1s : 1. Path to a chromsizes file, 2. Bin size in bp : Path to BED file defining the genomic bin segmentation. PIXELS_PATH : Text file containing nonzero pixel values. May be gzipped. Pass '-' to use stdin. COOL_PATH : Output COOL file path or URI. **Notes** Two input format options (tab-delimited). Input pixel file may be compressed. COO: COO-rdinate sparse matrix format (a.k.a. ijv triple). 3 columns: "bin1_id, bin2_id, count", BG2: 2D version of the bedGraph format. 7 columns: "chrom1, start1, end1, chrom2, start2, end2, count" **Examples** cooler load -f bg2 : in.bg2.gz out.cool NrZreflectrZdropbin1_idbin2_id)r r!countr)chrom1start1end1chrom2start2end2r")r#r$r%r&r'r(r"r#rr$rr%rr&r'r(T)Zschema is_one_based tril_actionsortr)r-r.r/F)Z includes_agg>r r!zA field number is required.)Z param_hint- csg|]}|qSr2).0nameZinput_field_numbersr2X/var/www/html/software/conda/envs/catlas/lib/python3.10/site-packages/cooler/cli/load.py Mszload..)sepZusecolsnamesdtypecommentiterator chunksizezfields: zdtypes: zsymmetric-upper: aw) columnsZdtypesmetadataassemblymergebuf max_mergetemp_dirZ delete_tempZ ensure_sortedZ triuchecksymmetric_upperh5optsmode)r __name__r openjsonloadrrstrintgetrrlenr appendclickZ BadParameternpZfloat64r isinstancelisttuplesysstdinpdZread_csvinformap)(rrrformatrArBfieldZcount_as_floatZ one_basedZ comment_charZinput_copy_statusZno_symmetric_upperr=rCrDrEZno_delete_tempZstorage_optionsrQkwargsloggerZ chromsizesZbinsrFr.fZoutput_field_namesZoutput_field_dtypesZinput_field_namesZinput_field_dtypesZpipelineargr4Zcolnumr:_rGkeyZf_inreaderr2r5r6rLs/                   "         rL)rWrRnumpyrSZpandasrYZ simplejsonrKcreaterrrrrr r Z_utilr r r commandargumentPathoptionZChoicerMrNrLr2r2r2r6s       4