B ž3Rcp?ã @sædZddlZddlZddlmZddlmZddlmZddlm Z ddl m Z m Z m Z ddlZdZd d „Zdejfd d „Zejfd d„Zdggdfdd„ZdTdd„ZdUdd„Zdd„ZedkrâddlZy e ejdd…dd¡\ZZWn,ek rddlZe ¡eƒYnXe   ¡Z e  !e ¡dge _"de _#de _$de _%de _&de _'de _(de _)de _*de _+de _,de _-x¾eD]´\Z.Z/e.dkr–e/e _0e.dkrªe/e _%q|e.dks¾e.d krèe1e/ƒe _"e2e j"e3e4fƒs0e j"fe _"q|e.d!kre5e/ƒe _6q|e.d"kre/e _7q|e.d#kr(e/e _8q|e.d$kr:de _9nöe.d%krRde _#de _$nÞe.d&krjde _#de _$nÆe.d'kr€e5e/ƒe _:n°e.d(kr–e5e/ƒe _;nše.d)kr¨e/e _&nˆe.d*krºde _'nve.d+krÔee j0e j%ej?ej@e j(ejAd3ZBeB C¡ZDeB E¡ZFed4eFƒe GeFd5fej5¡ZHe GeFejI¡ZJdgeFZKgZLe j8rØee j0e j8ƒZMed6ƒeM N¡ZOeO Pd7e j8e j7f¡ed8ƒy eO Q¡ZRWnek r2dZRYnXxœeRrÐed9eSeLƒƒeRdZRyeL Te UeVeRƒ¡¡Wn.ek ršddlZe ¡e=d:ƒYn Xed;ƒy eO Q¡ZRWnek rÊdZRYnXq6WdZOnbx`eD]XZWye XeYeWd<ƒ¡ZZWn2ek r(ddlZe=d=eWƒe ¡Yn XeL TeZ¡qÞWeSeLƒZ[iZ\d>geSeLƒZ]x@e^eLƒD]2\Z_ZZeBZ`y eZjaZbWneck rŽYn Xe  deb¡e j9r°e jee`dd?efeZd@ƒr0e j$sÌe j#r0e jge` E¡eZjhddA\ZiZje j:dBkre jke`e j;e j:deiddC\ZlZmejem7ZjelZie j#r@eiejZjZine4ene` E¡ƒƒZjedDeSejƒƒeeZeDe`e j"eje j)dE\ZoZpedFƒeepe j*e j6dGZqe j+r¨eYdHe j+e_dfdIƒZrndZrxÔe^eqƒD]È\ZsZteqesZtetdeKes<eHesdetd7<eHesdetd57<e\ uesg¡Zvev Tetd¡eve\es<eJesd7<errLer wdJetd¡etdeod5kr¶etd5e]e_kr¶etd5e]e_<q¶WedKe]e_ƒq`We j&r¸ee eHeJe\e[eodLn e[dkrÌe=dMƒne=dNƒdZsxêeseFkrÂeJesdkrÂe[dkrpe  xe\es¡\ZyZze j{ezeSe\esƒdOdPZ|e=dQesdeHesdeJese|eHesdeJeseJeseVeKesƒfƒnFe=dResdeHesdeJeseHesdeJeseJeseVeKesƒfƒesd7ZsqÚWe  xe]¡\ZyZze=dSeyezfƒdS)Va÷Command line tool to construct an enrichment plot from saved composite models Usage: EnrichPlot [optional args] -d dbname -t tablename 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. éN)Ú DataStructs)ÚRDConfig)Ú DbConnect)Ú CompositeRun)Ú DataUtilsÚ SplitDataÚStatsz2.4.0cCs||kdp||kdS)Néÿÿÿÿé©)Zt1Zt2r r ú2lib/python3.7/site-packages/rdkit/ML/EnrichPlot.pyÚcmpUsr cCs&|r| d|¡n| d|¡dS)z­ emits messages to _sys.stderr_ override this in modules which import this one to redirect output **Arguments** - msg: the string to be displayed z%s z%s N)Úwrite)ÚmsgZnoRetÚdestr r r ÚmessageYs rcCstj d|¡dS)z­ emits messages to _sys.stderr_ override this in modules which import this one to redirect output **Arguments** - msg: the string to be displayed z ERROR: %s N)ÚsysÚstderrr)rrr r r Úerrorhs rr cCs`| |¡x`tt|ƒƒD]P}| |¡}t|dƒrt|jtƒsi}t|dƒrbx|jD] } d|| <qRW||_qWg} | ¡r~d} nd} |s–t tt|ƒƒƒ}d} x¼|D]´} |ræg}xÓóz"AccumulateCounts..rr )Úsortrrr) Z predictionsÚthreshÚsortItr%ZnCorrectÚnPtsr'ZIDÚrealr(r)r r r ÚAccumulateCountsºsr5r cCst|dƒr|jsdSd|j}t|dƒ}d}xÚ|t|ƒkr||dkr|dkr¾t ||¡\} } tj| t||ƒdd} | d|d||d||||d||||| f¡n>| d |d||d||||d||||f¡|d7}q.W| ¡d |j} t| dƒ} d t }t || d |dkrRt d || d t d| d |dkr˜|d}t d|d| dt d||f| d nt d|| d |  ¡t|dƒr|j ry*ddl m }|ƒ}|d| ƒt dƒWn&tk rddl}| ¡YnXdS)NÚplotFilez%s.datzw+rr éZ)Úlevelz%d %f %f %d %f z %d %f %f %d z%s.gnuzõ# 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 )Úfilez set yr [0:%d]zplot x with lineséz!replot "%s" using 1:2 with lines,Ú)Úendr9z)"%s" every %d using 1:2:5 with yerrorbarszreplot "%s" with pointsÚshowPlot)ÚGnuplotz load "%s"zpress return to continue... )rr6ÚopenrrÚ MeanAndDevÚGetConfidenceIntervalrÚcloseÚ__VERSION_STRINGÚprintr=r>ÚinputÚ ExceptionÚ tracebackÚ print_exc)ÚdetailsÚfinalÚcountsÚ pickVectsÚnModelsÚ nTrueActsZ dataFileNameZoutFr'Ú_ÚsdÚ confIntervalZ plotFileNameZgnuFZgnuHdrZeveryGapr>ÚprGr r r ÚMakePlotâsJ  00          rScCstj t¡t d¡dS)z$ displays a usage message and exits r N)rrrÚ__doc__Úexitr r r r ÚUsages rVÚ__main__zd:t:a:N:p:cSTHF:v:) zthresh=z plotFile=r=z pickleCol=ZOOBZnoSortz pickBase=ÚdoROCz rocThresh=zenrich=r;z-dz-tz-az--enrichz--threshz-Nz-pz-Sz-Hz-Tz-Fz-vz --plotFilez --showPlotz --pickleColz--OOBz--noSortz--doROCz --rocThreshz --pickBasez2*******Please provide both the -d and -t argumentszBuilding Data set )ÚuserZpasswordÚ pickleColZ pickleClassz npts: %d éz"-> Retrieving models from databasez$select model from %s where note='%s'z-> Reconstructing modelsz Building model %dz Model failedz <-DoneÚrbzproblems with model %s:g„×—A)ZshuffleÚ _splitFrac)Zsilentg)Z indicesToUseZ indicesOnlyzscreening %d examples)r r!r"Z accumulating)r2r1z %s.%d.pickszw+z%s zHalfway point: %d )rNzH#Index Avg_num_correct Conf90Pct Avg_num_picked Num_picks last_selectionz>#Index Avg_num_correct Avg_num_picked Num_picks last_selectionr7)r8z%d %f %f %f %d %sz%d %f %f %d %szHalfway point: %.2f(%.2f))rr )r )}rTrZnumpyZrdkitrrZrdkit.Dbase.DbConnectionrZrdkit.MLrZ rdkit.ML.DatarrrÚpicklerCr rrrr*r5rSrVÚ__name__ZgetoptÚargvÚargsZextrasrFrGrHrIZ SetDefaultsZ activeTgtZ doTrainingZ doHoldoutZ dbTableNamer6r=rZr"r2ZpickBaserXZ rocThreshÚargÚvalZdbNameÚevalrÚtuplerÚfloatZ thresholdZnoteZpersistTblNameZshuffleActivitiesZ filterFracZ filterValÚintrDZDBToDataZ defaultDBUserZdefaultDBPasswordZExplicitBitVectZdataSetZ GetVarNamesrZGetNPtsr3ZzerosrJZintegerrKZselPtsZmodelsZconnZ GetCursorZcursZexecuteZfetchoneZblobrrÚloadsÚstrZ modelNameÚloadr?ZmodelrMrLZ halfwayPtsÚ enumerateZ whichModelZtmpDZ _randomSeedZseedÚAttributeErrorZInitRandomNumbersZRandomizeActivitiesrZ SplitIndicesr]ZtrainIdxZtestIdxZ FilterDataZ trainFiltZtemprr&Z screenResZ runningCountsZpickFiler'Úentryrr$rr@ZmeanrPrArQr r r r ÚBs~     F ( 6                                            $   ,.