&hZdZddlZddlmZejde ddlZejZ dZ ejjdZ eje ZGd d Zd Zd Zd Z d dZdZ d!dZdZdZdZ d"dZdZd#dZdZdZdZdZ dZ!dZ"dZ#dZ$dZ%y#e$rddlmZeddwxYw)$aA state-emitting MarkovModel. Note terminology similar to Manning and Schutze is used. Functions: train_bw Train a markov model using the Baum-Welch algorithm. train_visible Train a visible markov model using MLE. find_states Find the a state sequence that explains some observations. load Load a MarkovModel. save Save a MarkovModel. Classes: MarkovModel Holds the description of a markov model N)BiopythonDeprecationWarningzThe 'Bio.MarkovModel' module is deprecated and will be removed in a future release of Biopython. Consider using the hmmlearn package instead.)MissingPythonDependencyErrorzRPlease install NumPy if you want to use Bio.MarkovModel. See http://www.numpy.org/cli}t|ddd}t|dz }|D] \}}||z ||<|S)zAReturn a dictionary of values with their sequence offset as keys.N) enumeratelen)valuesdentriesnindexkeys ^/mounts/lovelace/software/anaconda3/envs/py312/lib/python3.12/site-packages/Bio/MarkovModel.py itemindexr1sK Att %G F aA sU# HgYnc eZdZdZ ddZdZy) MarkovModelz+Create a state-emitting MarkovModel object.NcJ||_||_||_||_||_y)zInitialize the class.N)statesalphabet p_initial p_transition p_emission)selfrrrrrs r__init__zMarkovModel.__init__Ds(   "($rcvddlm}|}t|||jd|j S)z9Create a string representation of the MarkovModel object.r)StringIO)iorsaveseekread)rrhandles r__str__zMarkovModel.__str__Ns- T6 A{{}rNNN)__name__ __module__ __qualname____doc__rr$rrrrAs5OS%rrcj|j}|j|std|d||S)zNRead the first line and evaluate that begisn with the correct start (PRIVATE).z I expected z but got )readline startswith ValueError)r#startlines r_readline_and_check_startr1Xs7 ?? D ??5 !;uiyABB Krc t|d}|jdd}t|d}|jdd}t||}t|t|}}t j ||_t|d}tt|D]>}t|d||d}t|jd|j |<@t j ||f|_ t|d }tt|D]P}t|d||d}|jddDcgc] }t|c}|j|ddf<Rt j ||f|_ t|d }tt|D]P}t|d||d}|jddDcgc] }t|c}|j|ddf<R|Scc}wcc}w) z.Parse a file handle into a MarkovModel object.zSTATES:rNz ALPHABET:zINITIAL: :rz TRANSITION:z EMISSION:) r1splitrr npzerosrrangefloatrr) r#r0rrmmNMivs rloadr?`s %VY 7D ZZ\!" F %V[ 9Dzz|ABH VX &B v;H qA88A;BL $VZ 8D 3v; 2(2fQi[1BC R 01 Q2 hh1v&BO $V] ;D 3v; E(2fQi[1BC37::<3C Daq D1E HHaV$BM $V[ 9D 3v; C(2fQi[1BC15ab1ABAuQxB adC I!ECs G<Hc |j}|ddj|jd|ddj|jd|dt t |j D],}|d|j|d|j |dd.|d t t |jD]A}|d|j|ddjd |j|DdC|d t t |jD]A}|d|j|ddjd |j|DdCy )z$Save MarkovModel object into handle.zSTATES:   z ALPHABET: z INITIAL: r3z: gz TRANSITION: c32K|]}t|ywNstr.0xs r zsave..s'K1A'Kz EMISSION: c32K|]}t|ywrErFrHs rrKzsave..s'I1A'IrLN) writejoinrrr8r rrr)r:r#wr=s rr r sW  A"))$%R() 388BKK(),-lO 3r||$ %6 Bryy|nBr||Aq1 456o 3r' (Q Bryy|nBsxx'K8J'KKLB OPQm 3r}}% &O Bryy|nBsxx'I a8H'IIJ" MNOrc t|t|}}|s td|0tj|}|j|fk7r td|1tj|}|j||fk7r td|1tj|}|j||fk7r tdg} t |} |D]$} | j | D cgc]} | |  c} &| D cgc] } t| } } t| dk(r tdt||| ||||} | \}}}t|||||Scc} wcc} w)a Train a MarkovModel using the Baum-Welch algorithm. Train a MarkovModel using the Baum-Welch algorithm. states is a list of strings that describe the names of each state. alphabet is a list of objects that indicate the allowed outputs. training_data is a list of observations. Each observation is a list of objects from the alphabet. pseudo_initial, pseudo_transition, and pseudo_emission are optional parameters that you can use to assign pseudo-counts to different matrices. They should be matrices of the appropriate size that contain numbers to add to each parameter matrix, before normalization. update_fn is an optional callback that takes parameters (iteration, log_likelihood). It is called once per iteration. zNo training data given.$pseudo_initial not shape len(states)5pseudo_transition not shape len(states) X len(states)5pseudo_emission not shape len(states) X len(alphabet)rz,I got training data with outputs of length 0)pseudo_initialpseudo_transitionpseudo_emission update_fn) r r.r6asarrayshaperappendmin _baum_welchr)rr training_datarUrVrWrXr;r<training_outputsindexesoutputsrJlengthsrrrs rtrain_bwrcsu4 v;H qA 233!N3   A4 'CD D$JJ'89  " "q!f ,TU U"**_5  QF *TU U !G ?W = =>? 00!s1v0G0 7|qGHH   %+' A+,'I|Z vxL* MM%!>1s  E 6Eic | t|}n t||f}|t||f}nt|||f}|t||f}nt|||f}tj|} tj|} tj|} |tj|} nd} |tj|}nd}|tj|}nd}d}t t D]Q}t }|D]}|t|||| | | | || z }| | |||tj||z dkrn|}Stdt z| | | fDcgc]}tj|c}Scc}w)zfImplement the Baum-Welch algorithm to evaluate unknown parameters in the MarkovModel object (PRIVATE).Ng?z%HMM did not converge in %d iterations) _random_norm_copy_and_checkr6logr8MAX_ITERATIONSLOG0_baum_welch_onefabs RuntimeErrorexp)r;r<r_rrrrUrVrWrX lp_initial lp_transition lp_emissionlpseudo_initiallpseudo_transitionlpseudo_emission prev_llikr=llikra_s rr]r]s O #It4 #QF+ &|aV< !1a&) $Z!Q8  "JFF<(M&&$K!&&0$VV$56!"66/2 I > "U' G O"   D   a   RWWY-=%>%D  %U(B^STT!+M;G H!BFF1I HH Hs#Fc @t|} t|| ||||} t|| |||} tj||| f} t | D]} || }tj||f}t |D]@}t |D]0}| || |||z|||z| || dzz}||||<2B|t |z | dddd| f<tj|| f}t | D]+} t |D]}t | |dd| f||| <-tj|}t |D]}t ||ddf||<|dddf}|t||}|t |z }t |D]]}t |D]!}t | ||ddf||z |||<#|5t|||||<||t ||z ||<_t |D]}tj|tz}t | D]/} || }t |D]}t||| ||| f||<1|t |z }|t|||}|t |z }|||ddf<t | dd| fS)zExecute one step for Baum-Welch algorithm (PRIVATE). Do one iteration of Baum-Welch based on a sequence of output. Changes the value for lp_initial, lp_transition and lp_emission in place. rNr) r _forward _backwardr6r7r8_logsum _logvecaddri logaddexp)r;r<rarnrorprqrrrsTfmatbmatlp_arctk lp_traverser=jlp lp_arcout_t lp_arcoutksums rrjrj)s/ G A Aq*m[' JD Q=+w ?DXXq!Qi F 1X= AJhh1v& q 'A1X ' GAJ#A&q)*!!nQ'(1ga!en% %' Aq! ' '& (<<q!Qw#=(((Aq6"K 1X9q 9A 'q!Qw 8KN1  99  I 1X2{1a401 ! 2QT"J" O< '*"55 1XLq JA")&Aq/":Yq\"IM! Q  J  ))-*:A A1X >#DGVAq!G_=Q > >gdm#  'd$4Q$78D'$-'D AqD !& 41: rc2tj||dzf}||dddf<td|dzD]a}||dz }t|D]I} t} t|D]+} || |dz || | z|| |z} t | | } -| || |<Kc|S)zImplement forward algorithm (PRIVATE). Calculate a Nx(T+1) matrix, where the last column is the total probability of the output. rNrr6r7r8rir|) r;r}rnrorpramatrixrrrlprobr=rs rrxrxs XXq!a%j !FF1a4L 1a!e_ ! AENq !AE1X -AYq1u% a(8(;;k!nQ>OO!%, -!F1IaL ! ! Mrctj||dzf}t|dz ddD]^}||}t|D]I}t} t|D]+} || |dz||| z|||z} t | | } -| |||<K`|S)z'Implement backward algorithm (PRIVATE).rrr) r;r}rorprarrrr=rrrs rryrys XXq!a%j !F 1q5"b ! ! AJq !AE1X -AYq1u% a(8(;;k!nQ>OO!%, -!F1IaL ! ! Mrc t|t|}}|0tj|}|j|fk7r t d|1tj|}|j||fk7r t d|1tj|}|j||fk7r t dgg} }t |} t |} |D]k\} } t| t| k7r t d|j | Dcgc]}| | c}| j | Dcgc]}| | c}mt||| ||||}|\}}}t|||||Scc}wcc}w)a Train a visible MarkovModel using maximum likelihoood estimates for each of the parameters. Train a visible MarkovModel using maximum likelihoood estimates for each of the parameters. states is a list of strings that describe the names of each state. alphabet is a list of objects that indicate the allowed outputs. training_data is a list of (outputs, observed states) where outputs is a list of the emission from the alphabet, and observed states is a list of states from states. pseudo_initial, pseudo_transition, and pseudo_emission are optional parameters that you can use to assign pseudo-counts to different matrices. They should be matrices of the appropriate size that contain numbers to add to each parameter matrix. rRrSrTzstates and outputs not aligned) r r6rYrZr.rr[_mler)rrr^rUrVrWr;r<training_statesr_states_indexesoutputs_indexestoutputststatesrJrrrs r train_visiblers. v;H qA!N3   A4 'CD D$JJ'89  " "q!f ,TU U"**_5  QF *TU U )+B%Ov&N)O*H' w<3x= (=> >7Caq 1CDX F!3 FG H    A+,'I|Z vxL* MM D Fs  E $ E$ c tj|}|r||z}|D]}||dxxdz cc<t|}tj||f} |r| |z} |D]:}tt |dz D]} || || dz} } | | | fxxdz cc< <tt | D]%} | | ddft | | ddfz | | ddf<'tj||f} |r| |z} tj ||f} t||D](\}}t||D]\}}| ||fxxdz cc<*tt | D]%} | | ddft | | ddfz | | ddf<'|| | fS)z>?M&&)::;K $G"( )Qgaj )F )q*m[&IG 3w< C ,23qbiil3RVVE]B C N*4s # D#Dct|}g}t|D]}|jdg|ztj||f}||dd|dfz|dddf<td|D]U} || } t|D]@} |dd| dz f|dd| fz|| | fz} t | } | | d|| | f<| || | <BWg}g}t |dd|dz f} | D]#}|j|dz |g|||dz f%|r]|j \} }}| dk(r|j||fn-||d| } | D]}|j| dz |g|z|f|r]|S)zSImplement Viterbi algorithm to find most likely states for a given input (PRIVATE).Nrr)r r8r[r6r7rpop)r;rnrorprr} backtracer=scoresrrri_scoresr` in_processrrrs rrr:s F AI 1X%$!$%XXq!f F AvayL 99F1a4L 1a[& 1Iq &AaQh'-1*== AqD@QQH)G#GAJ/F1a4L%IaLO  &&JGq!a%x()G :1q51#vayQ'789: %>>+65 6 NNFE? +q *1-G @!!1q51#,">? @  Nrct|jdk(r|t|z }|St|jdk(r>tt|D]%}||ddft||ddfz ||ddf<'|St d)z"Normalize matrix object (PRIVATE).rNz&I cannot handle matrixes of that shape)r rZrr8r.)rr=s rrrcs 6<<A#f+% M V\\ a s6{# 2 dimensions)r6arrayrZr.r rkrr8)r desired_shaper=s rrfrf|sXXf1 %F ||}$.// 6<<A 773v;$ % ,;< < M V\\ a s6{# HAwws6!9~+,t3 !BQ!FGG H MABBrct|jdkDr5tj|tj|jf}n|}t }|D]}t ||}|S)z/Implement logsum for a matrix object (PRIVATE).r)r rZr6reshapeprodrir|)rvecrnums rrzrzs] 6<<1jj"''&,,"7!9: C"S!" Jrct|t|k(sJdtjt|}tt|D]}t ||||||<|S)z5Implement a log sum for two vector objects (PRIVATE).zvectors aren't the same length)r r6r7r8r|)logvec1logvec2sumvecr=s rr{r{sb w<3w< 'I)II XXc'l #F 3w< 6gaj'!*5q 6 MrcBt|}tj|S)z-Return the exponential of a logsum (PRIVATE).)rzr6rm)numbersrs r _exp_logsumrs ' C 66#;r)NNNN)NNNNNNNr%rE)&r)warningsBiorwarnnumpyr6 ImportErrorrr|rrseedrrgrirr1r?r rcrhr]rjrxryrrrrrrrrerfrzr{rr*rrrsG"+  P LL    rvv ."JO,BNJGIT[|. ( <N~+/\ 4&R   ( 0 & $  s B==C