ó a#nSc @s( dZddlmZddlZddlZddlZddlmZmZm Z ddl m Z ddlm Z ddl mZddlZddlZddlZdZd ejd „Zejd „Zd ggd d „Zd d d„Zdd„Zd„Zedkr$ ddlZy&ejejd dd\ƒ\ZZWn$ddlZej ƒeƒnXejƒZ!ej"e!ƒd ge!_#d e!_$d e!_%de!_&de!_'d e!_(de!_)d e!_*d e!_+de!_,d e!_-de!_.xUeD]M\Z/Z0e/dkr"e0e!_1ne/dkr:e0e!_&qþe/d ksRe/d!kr—e2e0ƒe!_#e3e!j#ƒej4ej5fkrKe!j#fe!_#qKqþe/d"krµe6e0ƒe!_7qþe/d#krÍe0e!_8qþe/d$kråe0e!_9qþe/d%krýd e!_:qþe/d&krd e!_$d e!_%qþe/d'kr?d e!_$d e!_%qþe/d(kr]e6e0ƒe!_;qþe/d)kr{e6e0ƒe!_<qþe/d*kr“e0e!_'qþe/d+kr«d e!_(qþe/d,krÍe=e0ƒd e!_)qþe/d-kråd e!_*qþe/d.krýd e!_+qþe/d/krd e!_-qþe/d0kr3e=e0ƒe!_.qþe/d1krþe0e!_,qþqþWe!j1 sce!j& rreƒd2GHned3ƒej>e!j1e!j&d4ej?d5ej@d6e!j)d7e jAƒZBeBjCƒZDeBjEƒZFed8eFƒejGeFd9fej6ƒZHejGeFejIƒZJdgeFZLgZMe!j9r?e e!j1e!j9ƒZNed:ƒeNjOƒZPePjQd;e!j9e!j8fƒed<ƒyePjRƒZSWn dZSnXx—eSr5ed=eTeMƒƒeSd ZSy eMjUejVeWeSƒƒƒWn"ddlZej ƒd>GHn Xed?ƒyePjRƒZSWqŸdZSqŸXqŸWdZPncx`eD]XZXyejYeZeXd@ƒƒZ[Wn&ddlZdAeXGHej ƒqFXeMjUe[ƒqFWeTeMƒZ\iZ]dBgeTeMƒZ^xÞe_eMƒD]Ð\Z`Z[eBZay e[jbZcWnedk rnXejeecƒe!j:r3ejfeadCd ƒnege[dDƒrée!j%sTe!j$réejheajEƒe[jidEd ƒ\ZjZke!j;dFkrÍejleae!j<e!j;ddGejdHd ƒ\ZmZneken7ZkemZjne!j$rûejekZkZjqûneoeajEƒƒZkedIeTekƒƒee[eDeadJe!j#dKekdLe!j*ƒ\ZpZqedMƒeeqdNe!j+dOe!j7ƒZre!j,r“eZdPe!j,e`d fdQƒZsndZsxöe_erƒD]è\ZtZueretZueud eLet Required Arguments: -d "dbName": the name of the database for screening -t "tablename": provide the name of the table with the data to be screened : file name(s) of pickled composite model(s). If the -p argument is also provided (see below), this argument is ignored. Optional Arguments: - -a "list": the list of result codes to be considered active. This will be eval'ed, so be sure that it evaluates as a list or sequence of integers. For example, -a "[1,2]" will consider activity values 1 and 2 to be active - --enrich "list": identical to the -a argument above. - --thresh: sets a threshold for the plot. If the confidence falls below this value, picking will be terminated - -H: screen only the hold out set (works only if a version of BuildComposite more recent than 1.2.2 was used). - -T: screen only the training set (works only if a version of BuildComposite more recent than 1.2.2 was used). - -S: shuffle activity values before screening - -R: randomize activity values before screening - -F *filter frac*: filters the data before training to change the distribution of activity values in the training set. *filter frac* is the fraction of the training set that should have the target value. **See note in BuildComposite help about data filtering** - -v *filter value*: filters the data before training to change the distribution of activity values in the training set. *filter value* is the target value to use in filtering. **See note in BuildComposite help about data filtering** - -p "tableName": provides the name of a db table containing the models to be screened. If you use this argument, you should also use the -N argument (below) to specify a note value. - -N "note": provides a note to be used to pull models from a db table. - --plotFile "filename": writes the data to an output text file (filename.dat) and creates a gnuplot input file (filename.gnu) to plot it - --showPlot: causes the gnuplot plot constructed using --plotFile to be displayed in gnuplot. iÿÿÿÿ(tRDConfigN(t DataUtilst SplitDatatStats(t DbConnect(t DataStructs(t CompositeRuns2.4.0icCs/|r|jd|ƒn|jd|ƒdS(s¥ emits messages to _sys.stderr_ override this in modules which import this one to redirect output **Arguments** - msg: the string to be displayed s%s s%s N(twrite(tmsgtnoRettdest((sU/public2/ftp/pub/ccp4/6.5/ccp4-6.5/lib/python2.7/site-packages/rdkit/ML/EnrichPlot.pytmessageOs cCstjjd|ƒdS(s¥ emits messages to _sys.stderr_ override this in modules which import this one to redirect output **Arguments** - msg: the string to be displayed s ERROR: %s N(tsyststderrR(RR ((sU/public2/ftp/pub/ccp4/6.5/ccp4-6.5/lib/python2.7/site-packages/rdkit/ML/EnrichPlot.pyterror\s icCsò|j|ƒx’tt|ƒƒD]~}|j|ƒ}t|dƒr t|jƒtjkr i}t|dƒr’x|jD]} d|| | |kr|d7}n|d7}|j|||fƒq>q>W|S(sú Accumulates the data for the enrichment plot for a single model **Arguments** - predictions: a list of 3-tuples (as returned by _ScreenModels_) - thresh: a threshold for the confidence level. Anything below this threshold will not be considered - sortIt: toggles sorting on confidence levels **Returns** - a list of 3-tuples: - the id of the active picked here - num actives found so far - number of picks made so far cSst|d|dƒS(Ni(tcmp(txty((sU/public2/ftp/pub/ccp4/6.5/ccp4-6.5/lib/python2.7/site-packages/rdkit/ML/EnrichPlot.pytÃsii(tsortRRR( t predictionstthreshtsortItR)tnCorrecttnPtsR,tidtrealR/R0((sU/public2/ftp/pub/ccp4/6.5/ccp4-6.5/lib/python2.7/site-packages/rdkit/ML/EnrichPlot.pytAccumulateCountsªs    cCs~t|dƒ s|j rdSd|j}t|dƒ}d}x|t|ƒkrI||dkrI|dkr÷tj||ƒ\} } tj| t||ƒddƒ} |jd|d||d||||d||||| fƒnE|jd |d||d||||d||||fƒ|d7}qCW|jƒd |j} t| dƒ} d t }| |IJ|dkr¡| d |IJn| d IJ|dkræ|d}| d|I| d||fIJn| d|IJ| jƒt|dƒrz|j rzyAddl }ddl m }|ƒ}|d| ƒt dƒWqzddl}|jƒqzXndS(NtplotFiles%s.datsw+iitleveliZs%d %f %f %d %f s %d %f %f %d s%s.gnusõ# Generated by EnrichPlot.py version: %s set size square 0.7 set xr [0:] set data styl points set ylab 'Num Correct Picks' set xlab 'Num Picks' set grid set nokey set term postscript enh color solid "Helvetica" 16 set term X s set yr [0:%d]splot x with linesis!replot "%s" using 1:2 with lines,s)"%s" every %d using 1:2:5 with yerrorbarssreplot "%s" with pointstshowPlotiÿÿÿÿ(tGnuplots load "%s"spress return to continue... (RR@topenRRt MeanAndDevtGetConfidenceIntervalRtcloset__VERSION_STRINGRBtosRCt raw_inputt tracebackt print_exc(tdetailstfinaltcountst pickVectstnModelst nTrueActst dataFileNametoutFR,tmeantsdt confIntervalt plotFileNametgnuFtgnuHdrteveryGapRIRCtpRK((sU/public2/ftp/pub/ccp4/6.5/ccp4-6.5/lib/python2.7/site-packages/rdkit/ML/EnrichPlot.pytMakePlotÑsN % "#(#"              cCs!tjjtƒtjdƒdS(s$ displays a usage message and exits iÿÿÿÿN(R R Rt__doc__texit(((sU/public2/ftp/pub/ccp4/6.5/ccp4-6.5/lib/python2.7/site-packages/rdkit/ML/EnrichPlot.pytUsage st__main__sd:t:a:N:p:cSTHF:v:sthresh=s plotFile=RBs pickleCol=tOOBtnoSorts pickBase=tdoROCs rocThresh=senrich=ts-ds-ts-as--enrichs--threshs-Ns-ps-Ss-Hs-Ts-Fs-vs --plotFiles --showPlots --pickleCols--OOBs--noSorts--doROCs --rocThreshs --pickBases2*******Please provide both the -d and -t argumentssBuilding Data set tusertpasswordt pickleColt pickleClasss npts: %d is"-> Retrieving models from databases$select model from %s where note='%s's-> Reconstructing modelss Building model %ds Model faileds <-Donetrbsproblems with model %s:g„×—Atshufflet _splitFractsilentgt indicesToUset indicesOnlysscreening %d examplesR"R#R$t accumulatingR:R9s %s.%d.pickssw+s%s sHalfway point: %d RRsH#Index Avg_num_correct Conf90Pct Avg_num_picked Num_picks last_selections>#Index Avg_num_correct Avg_num_picked Num_picks last_selectionRAiZs%d %f %f %f %d %ss%d %f %f %d %ssHalfway point: %.2f(%.2f)( sthresh=s plotFile=sshowPlots pickleCol=RbRcs pickBase=sdoROCs rocThresh=senrich=(~R^trdkitRtnumpytcPickletcopyt rdkit.ML.DataRRRtrdkit.Dbase.DbConnectionRRtrdkit.MLRR RIRRHR R RR2R?R]R`t__name__tgetopttargvtargstextrasRKRLRMt SetDefaultst activeTgtt doTrainingt doHoldoutt dbTableNameR@RBRhR$R:tpickBaseRdt rocThreshtargtvaltdbNametevalRt TupleTypetListTypetfloatt thresholdtnotetpersistTblNametshuffleActivitiest filterFract filterValtinttDBToDatat defaultDBUsertdefaultDBPasswordtExplicitBitVecttdataSett GetVarNamesR tGetNPtsR<tzerosRNtintegerRORtselPtstmodelstconnt GetCursortcurstexecutetfetchonetblobRRtloadststrt modelNametloadRDtmodelRQRPt halfwayPtst enumeratet whichModelttmpDt _randomSeedtseedtAttributeErrortInitRandomNumberstRandomizeActivitiesRt SplitIndicesRlttrainIdxttestIdxt FilterDatat trainFiltttempRR+t screenRest runningCountstpickFileR,tentryRR(RRERURVRFRW(((sU/public2/ftp/pub/ccp4/6.5/ccp4-6.5/lib/python2.7/site-packages/rdkit/ML/EnrichPlot.pytCs” $  C' ;                        !                                                           !         #  (   "