DUfddlZddlZddlmZmZmZddlm Z m Z ddl m Z gdZ d"dZd"dZd"d Zd"d Zd#d Zd$dZejejejejejejejejejejejejejejejejejej ejej!ej"ejej#ej$iZ%dZ& d%dZ' d&dZ(d'dZ) d(dZ* d)dZ+ d*dZ, d+dZ-d,dZ. d-dZ/ d.dZ0d/dZ1 d0dZ2 d1d!Z3dS)2N)arropschecks construction)_get_default_colnames_verify_columns) parse_region)select select_maskselect_indices select_labelsexpandoverlapclustermergecoverageclosestsubtractsetdiffcount_overlapstrim complement sort_bedframe assign_viewc|tn|\}}}t||||gt|\}}}|td| |||k} nT| tj}|||k|||k|||kz||||k|||kzzz} | S)a& Return boolean mask for all genomic intervals that overlap a query range. Parameters ---------- df : pandas.DataFrame region : str or tuple The genomic region to select from the dataframe in UCSC-style genomic region string, or triple (chrom, start, end). cols : (str, str, str) or None The names of columns containing the chromosome, start and end of the genomic intervals. The default values are 'chrom', 'start', 'end'. Returns ------- Boolean array of shape (len(df),) Nz*no chromosome detected, check region input)rrr ValueErrornpinfto_numpy) dfregioncolsckskekchromstartendmasks I/var/www/html/software/conda/lib/python3.11/site-packages/bioframe/ops.pyr r s(-1L&(((dJBBBR %%%$V,,E5# }EFFF }"v ;&C2%fslr"v~ .2"R& RVu_5 7  ==??cTtjt|||dS)a Return integer indices of all genomic intervals that overlap a query range. Parameters ---------- df : pandas.DataFrame region : str or tuple The genomic region to select from the dataframe in UCSC-style genomic region string, or triple (chrom, start, end). cols : (str, str, str) or None The names of columns containing the chromosome, start and end of the genomic intervals. The default values are 'chrom', 'start', 'end'. Returns ------- 1D array of int r)rnonzeror r r!r"s r*r r Es%( :k"fd33 4 4Q 77r+c:|jt|||S)a Return pandas Index labels of all genomic intervals that overlap a query range. Parameters ---------- df : pandas.DataFrame region : str or tuple The genomic region to select from the dataframe in UCSC-style genomic region string, or triple (chrom, start, end). cols : (str, str, str) or None The names of columns containing the chromosome, start and end of the genomic intervals. The default values are 'chrom', 'start', 'end'. Returns ------- pandas.Index )indexr r.s r*r r \s* 8KFD11 22r+c:|jt|||S)a Return all genomic intervals in a dataframe that overlap a genomic region. Parameters ---------- df : pandas.DataFrame region : str or tuple The genomic region to select from the dataframe in UCSC-style genomic region string, or triple (chrom, start, end). cols : (str, str, str) or None The names of columns containing the chromosome, start and end of the genomic intervals. The default values are 'chrom', 'start', 'end'. Returns ------- df : pandas.DataFrame Notes ----- See :func:`.core.stringops.parse_region()` for more information on region formatting. See also -------- :func:`select_mask` :func:`select_indices` :func:`select_labels` )locr r.s r*r r ts> 6+b&$// 00r+bothc|tn|\}}}tj|d|||g||td|I|dkrtdd|dz z||j||jz z}|j||g} n8|'t |tstd |}ntd |} |d ks|d kr||j|z | |<|d ks|d kr|||z| |<||dkr||jd||j||jz z tj z} tj | |j| | |<tj | |j| | |<|C| ||g| ||g<| ||g | | ||g<| S)a@ Expand each interval by an amount specified with `pad`. Negative values for pad shrink the interval, up to the midpoint. Multiplicative rescaling of intervals enabled with scale. Only one of pad or scale can be provided. Often followed by :func:`trim()`. Parameters ---------- df : pandas.DataFrame pad : int, optional The amount by which the intervals are additively expanded *on each side*. Negative values for pad shrink intervals, but not beyond the interval midpoint. Either `pad` or `scale` must be supplied. scale : float, optional The factor by which to scale intervals multiplicatively on each side, e.g ``scale=2`` doubles each interval, ``scale=0`` returns midpoints, and ``scale=1`` returns original intervals. Default False. Either `pad` or `scale` must be supplied. side : str, optional Which side to expand, possible values are 'left', 'right' and 'both'. Default 'both'. cols : (str, str, str) or None The names of columns containing the chromosome, start and end of the genomic intervals. Default values are 'chrom', 'start', 'end'. Returns ------- df_expanded : pandas.DataFrame Notes ----- See :func:`bioframe.trim` for trimming interals after expansion. NT raise_errorsr"z(only one of pad or scale can be suppliedrz multiplicative scale must be >=0g?rzadditive pad must be integerz$either pad or scale must be suppliedr3leftright)rr is_bedframervaluesdtypes isinstanceintcopyastyperint64minimummaximumround) r padscalesider"r#r$r%padstypes df_expandedmidss r*rrsR-1L&(((dJBB rBB<@@@@ S_CDDD   199?@@ @eai BrFMBrFM$AB 2r(# #s## =;<< <?@@@''))K v~~R&-$. B v~~R&4- B  77b6=C2b6=2b6=+H$I#Q#Q$$D!jR)?FFKO jR)?FFKO  +RH 5 ; ; = = RH +RH 5 < z/_overlap_intidxs...js$===4aaag===r+rV)rW idxs_pairs r*rYz$_overlap_intidxs..is6%>=9===r+rshapedtype)rr reset_indexgroupbyindicessetunionr:rarraysizeroverlap_intervalswherebincountlen ones_likeappendblockndarrayr=vstack)df1df2howcols1cols2onck1sk1ek1ck2sk2ek2 group_list1 group_list2 df1_groups df2_groups all_groupsstarts1ends1starts2ends2overlap_intidxs group_keysdf1_group_idxsdf2_group_idxsoverlap_intidxs_subboth_groups_nonemptyoverlap_idxs_locno_overlap_ids1no_overlap_ids2s r*_overlap_intidxsrs,D05})+++%MCc/4})+++%MCcC#sC)))C#sC))) //t/ $ $C //t/ $ $C%K%K ~r r [4FFNJ[4FFNJ3z??C OO<QTU>U  %7'n%'n%     "#3AAAqD#9:"#3AAAqD#9:$   # # #(;a(?(?# 1"0H ,QQQT2c.>Q>Q  ##1 #o666$   $ $ $)Q>Q  ##1 o666#$      " ")<    ?q  zc2222i00O r+c tj|}n#t$r|cYSwxYwt||SrU)rr_ TypeErrorNUMPY_INT_TO_DTYPEgetr_s r*_to_nullable_dtypersP    ! !% / //s  &&TF_c J |tn|\} } }| tn| \}}}tj|d| | |gtj|d|||g|dkr|d}|dkr|rtd| [t | t std| | vs|| vrtdt || t || t||||| | }|ddd f}|ddd f}t |tr|nd }|d z}|d z}tj ||j |itj }tj ||j |itj }d}|rt |tr|nd }|dz| z}|dz|z}tj|| j|||j|}tj||j|||j|}tj ||||i}d} d}!|st|dks|dkr9|j|d} fd| jD| _|st|dks|dkr9|j|d}!fd|!jD|!_|dkr2|dk}"|dk}#|"}$|#}%d||"<d||#<| f| r]|$r[| | d zt-|| j|d zt-||ji} |$rd| |"<|!f| r]|%r[|!|d zt-||j|d zt-||ji}!|%rd|!|#<|"| r |$s|%r|}d||"|#z<tj|| ||!|gd}&|r|&||g}&|s|&||gd d|&dd|&S)an Find pairs of overlapping genomic intervals. Parameters ---------- df1, df2 : pandas.DataFrame Two sets of genomic intervals stored as a DataFrame. how : {'left', 'right', 'outer', 'inner'}, default 'left' How to handle the overlaps on the two dataframes. left: use the set of intervals in df1 right: use the set of intervals in df2 outer: use the union of the set of intervals from df1 and df2 inner: use intersection of the set of intervals from df1 and df2 return_input : bool, optional If True, return columns from input dfs. Default True. return_index : bool, optional If True, return indicies of overlapping pairs as two new columns ('index'+suffixes[0] and 'index'+suffixes[1]). Default False. return_overlap : bool, optional If True, return overlapping intervals for the overlapping pairs as two additional columns (`overlap_start`, `overlap_end`). When `cols1` is modified, `start` and `end` are replaced accordingly. When `return_overlap` is a string, its value is used for naming the overlap columns: `return_overlap + "_start"`, `return_overlap + "_end"`. Default False. suffixes : (str, str), optional The suffixes for the columns of the two overlapped sets. keep_order : bool, optional If True and how='left', sort the output dataframe to preserve the order of the intervals in df1. Cannot be used with how='right'/'outer'/'inner'. Default True for how='left', and None otherwise. Note that it relies on sorting of index in the original dataframes, and will reorder the output by index. cols1, cols2 : (str, str, str) or None, optional The names of columns containing the chromosome, start and end of the genomic intervals, provided separately for each set. The default values are 'chrom', 'start', 'end'. on : list or None, optional List of additional shared columns to consider as separate groups when considering overlaps. A common use would be passing on=['strand']. Default is None. ensure_int : bool, optional [default: True] If True, ensures that the output dataframe uses integer dtypes for start and end coordinates. This may involve converting coordinate columns to nullable types in outer joins. Default True. Returns ------- df_overlap : pandas.DataFrame Notes ----- If ``ensure_int`` is False, inner joins will preserve coordinate dtypes from the input dataframes, but outer joins will be subject to native type casting rules if missing data is introduced. For example, if `df1` uses a NumPy integer dtype for `start` and/or `end`, the output dataframe will use the same dtype after an inner join, but, due to casting rules, may produce ``float64`` after a left/right/outer join with missing data stored as ``NaN``. On the other hand, if `df1` uses Pandas nullable dtypes, the corresponding coordinate columns will preserve the same dtype in the output, with missing data stored as ``NA``. NTr5r7z+keep_order=True only allowed for how='left'on=[] must be None or list,on=[] should not contain chromosome colnamesrrrsrtrurrr0rrr1rLc&g|] }|dzSrrVrWcsuffixess r*rYzoverlap..!JJJ!a(1+oJJJr+2r8c&g|] }|dzSrrVrs r*rYzoverlap.. rr+innerrRcolumnsaxis)rinplacerMr)rrr9rr<listrrstrpd DataFramer0 Int64DtyperrBr:rAilocr`ranyr?rr_convert_dtypesconcat sort_valuesrM)'rprqrr return_input return_indexreturn_overlapr keep_orderrsrtru ensure_intrvrwrxryrzr{overlap_df_idxsevents1events2 index_col index_col_1 index_col_2 df_index_1 df_index_2 df_overlap overlap_coloverlap_col_sk1overlap_col_ek1 overlap_start overlap_end df_input_1 df_input_2 is_na_left is_na_right any_na_left any_na_rightout_dfs' ` r*rrsyj05})+++%MCc/4})+++%MCc sS#sODDDD sS#sODDDD v J.  v : FGGG ~"d## ;9:: : 2II3"99KLL LR   R   &   Oaaad#Gaaad#G!+< = =J 7Ihqk)Khqk)K{CIg,>?r}WWWJ{CIg,>?r}WWWJJ (2>3(G(GVnnY %+c1%+c1 HOG $ HOG $  j HOG $ HOG $  \ m_k J  JJKs<((C//<63I3IXg&222== JJJJz7IJJJ Ks<((C//<73J3JXg&222== JJJJz7IJJJ  g~~] m  nn&& "(( !% :"& ;  ! k '..hqk)+=c#hn+M+Mhqk)+=c#hn+M+M  .)- :&  ! l '..hqk)+=c#hn+M+Mhqk)+=c#hn+M+M  /*. ;'  ! <{ ?? F [+.Q EEE D$/// Mr+c . ||dkrtd|tn|\t|g|j}|d}g}|Lt |t std|vrtdt||||z }||dj} tj |j dd } g} d } | D]\} }tjtj| r?|jrG|j|}t'j|jtj|jtj| \}}}tj|}|| d zz }|j d}| |z } || |j<t | t2r| f} i}|D]R}tjtj || ||||j ||<S||<||<||d <tj|}| |tj|g|d }|dkrY| d ztjtj|jz| |j<| |j|tj | d} |dkr;| tj!tj!i} tj"| dksJt | #}| gfd|Dz} i}|r| |d<|r,| j| |d<| j| |d<tj|}|rtj ||gd}|$||S)a Cluster overlapping intervals into groups. Can return numeric ids for these groups (with `return_cluster_ids`=True) and/or their genomic coordinates (with `return_cluster_intervals`=True). Also see :func:`merge()`, which discards original intervals and returns a new set. Parameters ---------- df : pandas.DataFrame min_dist : float or None If provided, cluster intervals separated by this distance or less. If ``None``, do not cluster non-overlapping intervals. Since bioframe uses semi-open intervals, interval pairs [0,1) and [1,2) do not overlap, but are separated by a distance of 0. Such adjacent intervals are not clustered when ``min_dist=None``, but are clustered when ``min_dist=0``. cols : (str, str, str) or None The names of columns containing the chromosome, start and end of the genomic intervals. The default values are 'chrom', 'start', 'end'. on : None or list List of column names to perform clustering on independently, passed as an argument to df.groupby before clustering. Default is ``None``. An example useage would be to pass ``on=['strand']``. return_input : bool If True, return input return_cluster_ids : bool If True, return ids for clusters return_cluster_invervals : bool If True, return clustered interval the original interval belongs to Returns ------- df_clustered : pd.DataFrame Nrmin_dist>=0 currently requiredTrLrrrNrRmin_distrdatar_ n_intervalsrc"g|] }|fv | SrVrVrWcolr#r%r$s r*rYzcluster..)QQQRRL9P9P9P9P9Pr+r cluster_start cluster_endr)%rrrr0r`r<rragroupsrfullr^itemsrisnaSeriesremptyr2rmerge_intervalsr:r?r@rirr_rrlisnullsumarangerrallkeys set_index)r rr"rurreturn_cluster_idsreturn_cluster_intervalsdf_index group_list df_groups cluster_idsclustersmax_cluster_idr df_group_idxsdf_groupcluster_ids_groupcluster_starts_groupcluster_ends_groupinterval_counts n_clustersclusters_grouprdf_nansclusters_namesrr#r%r$s @@@r*rrSsh a<<=>> >,0L&(((dJBBBR %%%xH T " "BJ ~"d## ;9:: : 88KLL LBb  : 55>G{{}}q Q 26'.+A+A!B!B B GN# w(((y""..D.99H{{}}q??B R]__#MNN 6+" # ####(--//**N R QQQQQQ~QQQQHF('yA"*2,"5k"B ( 3K @} \& ! !F9B.hrr+)&rrrr9r>r`r<rrrarrrrrrrr2rrr:r?rr@rir^rrr0r_rrlrrNArrr)r rr"rurrrrrrrrrrrrrr df_has_nans nan_intervalsrr#r%r$s @@@r*rrsR a<<=>> >-1L&(((dJBB rBB<@@@@ BNN4dN+++J ~"d## ;9:: : 88KLL LBb  : 55>G++--K   UGk !"O&/'    I-0      y""..D.99H ?? "bmoo}bmoo V   (--//**N R QQQQQQ~QQQQH Or+c l|tn|\}}}|tn|\} } } |ddt||} t|| d|ddd|| } | d|| d|z | d<t j| d |d zdd id||j  dd i d}|rt j ||gd}|S)a Quantify the coverage of intervals from 'df1' by intervals from 'df2'. For every interval in 'df1' find the number of base pairs covered by intervals in 'df2'. Note this only quantifies whether a basepair in 'df1' was covered, as 'df2' is merged before calculating coverage. Parameters ---------- df1, df2 : pandas.DataFrame Two sets of genomic intervals stored as a DataFrame. suffixes : (str, str) The suffixes for the columns of the two overlapped sets. return_input : bool If True, return input as well as computed coverage cols1, cols2 : (str, str, str) or None The names of columns containing the chromosome, start and end of the genomic intervals, provided separately for each set. The default values are 'chrom', 'start', 'end'. Returns ------- df_coverage : pandas.DataFrame Notes ------ Resets index. NTrr"r7)rrrrrrrsrtoverlap_rr0rrrrrLrr) rr`rrrrraaggr?r_renamer)rprqrrrsrtrvrwrxryrzr{ df2_mergedrrs r*rrnshR05})+++%MCc/4})+++%MCcOODtO,,,s'''J     J''7#'7'78:FVQTFVFV;WWJy   w!4 5 5 S)U# $ $Y 0 VCHN # #  J/ 0 0 $   :C=y999 Mr+c |tn|\} } } | tn| \} }}d}|||ur&t|dkrtd|}d}|d}|d}|| dj}|| dj}g}|D]M\}}||vrFtj|dtj |zgj }| |P||}|j |}|j |}d}t|tr||j}n/t!|r||j}ntd d}|.tjt|tj }n||jd k}t'j|| j|| j|rdn ||j|rdn ||j|||||| }tjtjt||ddd fd}tjtj|j|ddd f|j|dddfgj tj|j|dtj |j|zgj g}| |Ot|d krtjdt2Stj|}|S)a  For every interval in set 1 find k closest genomic intervals in set2 and return their integer indices. Parameters ---------- df1, df2 : pandas.DataFrame Two sets of genomic intervals stored as a DataFrame. If df2 is None or same object as df1, find closest intervals within the same set. k_closest : int The number of closest intervals to report. cols1, cols2 : (str, str, str) The names of columns containing the chromosome, start and end of the genomic intervals, provided separately for each set. The default values are 'chrom', 'start', 'end'. Returns ------- closest_ids : numpy.ndarray The indices of the overlapping genomic intervals in the original dataframes. The 1st column contains the indices of intervals from the 1st set, the 2nd column - the indicies from the 2nd set. The second column is filled with -1 for those intervals in the 1st set with no closest 2nd set interval. NFrzKdf1 must have more than one interval to find closest non-identical intervalTrLrrRzHtie_breaking_col must be either a column label or f(DataFrame) -> Seriesr-)ktie_arrignore_overlapsignore_upstreamignore_downstream directionr)invertr[r])rrjrr`rarrrrorkTrlr2r<rr:callableonesbool_rclosest_intervalsisinr concatenaternr=)rprqr rrr direction_coltie_breaking_colrsrtrvrwrxryrzr{ self_closestr~rclosest_intidxsrrclosest_idxs_groupr df1_group df2_groupr  direction_arrna_idxss r*_closest_intidxsr#sT05})+++%MCc/4})+++%MCcL  s88q==)   //t/ $ $C //t/ $ $CS4007JS4007JO&0&6&6&8&8M3M3" N Z ' '!#"n555""    " "#5 6 6 6 #J/GN+ GN+  & , ,  018GG & ' ' &&y118GG )     GC NN"(CCCMM-(/36  $5 cN ! cN ! ;DDin&; ;DDin&;++/#    ' Ic.)) * *,>qqq!t,DT    ^ &-.@A.FG&-.@A.FG  &-g6R\.*?*HIII    " 12222 ?q  zc2222i00O r+c  |dkrtd||urtd||}| tn| \}}}|tn|\}}}tj|d|||gtj|d|||gt ||||||||| | }|dddfdk}d}d}| rt | t r| nd }tj| d z|j |ddd fi}tj| dz|j |dddfi}tj ||<d}| ratj tj ||j|ddd f||j|dddfgd }tjtj ||j|ddd f||j|dddfgd }||k}tj|tj||dtj||dd }|tjtjtjd }tj ||<d}| rtjd ||j|ddd f||j|dddfz }tjd ||j|dddf||j|ddd fz } tj tj || gd }!tjd |!itj}tj ||<d}"d}#|st |dks|dkrC|j|ddd fd}" fd|"jD|"_|st |dks|dkr|j|dddfd}# fd|#jD|#_|#| dztj| dztji}#tj |#|<tj||"||#||gd }$|$S)aY For every interval in dataframe `df1` find k closest genomic intervals in dataframe `df2`. Currently, we are not taking the feature strands into account for filtering. However, the strand can be used for definition of upstream/downstream of the feature (direction). Note that, unless specified otherwise, overlapping intervals are considered as closest. When multiple intervals are located at the same distance, the ones with the lowest index in `df2` are returned. Parameters ---------- df1, df2 : pandas.DataFrame Two sets of genomic intervals stored as a DataFrame. If `df2` is None, find closest non-identical intervals within the same set. k : int The number of the closest intervals to report. ignore_overlaps : bool If True, ignore overlapping intervals and return the closest non-overlapping interval. ignore_upstream : bool If True, ignore intervals in `df2` that are upstream of intervals in `df1`, relative to the reference strand or the strand specified by direction_col. ignore_downstream : bool If True, ignore intervals in `df2` that are downstream of intervals in `df1`, relative to the reference strand or the strand specified by direction_col. direction_col : str Name of direction column that will set upstream/downstream orientation for each feature. The column should contain bioframe-compliant strand values ("+", "-", "."). tie_breaking_col : str A column in `df2` to use for breaking ties when multiple intervals are located at the same distance. Intervals with *lower* values will be selected. return_input : bool If True, return input return_index : bool If True, return indices return_distance : bool If True, return distances. Returns zero for overlaps. return_overlap : bool If True, return columns: 'have_overlap', 'overlap_start', and 'overlap_end'. Fills df_closest['overlap_start'] and df['overlap_end'] with None if non-overlapping. Default False. suffixes : (str, str) The suffixes for the columns of the two sets. cols1, cols2 : (str, str, str) or None The names of columns containing the chromosome, start and end of the genomic intervals, provided separately for each set. The default values are 'chrom', 'start', 'end'. Returns ------- df_closest : pandas.DataFrame If no intervals found, returns none. Notes ----- By default, direction is defined by the reference genome: everything with smaller coordinate is considered upstream, everything with larger coordinate is considered downstream. If ``direction_col`` is provided, upstream/downstream are relative to the direction column in ``df1``, i.e. features marked "+" and "." strand will define upstream and downstream as above, while features marked "-" have upstream and downstream reversed: smaller coordinates are downstream and larger coordinates are upstream. rz k>=1 requiredzJpass df2=None to find closest non-identical intervals within the same set.NTr5)r rrrrrrsrtrRr0rr) have_overlaprrdistancerrr7rLc&g|] }|dzSrrVrs r*rYzclosest..rr+rr8c&g|] }|dzSrrVrs r*rYzclosest.."rr+r)rrrr9r#r<rrrr0rramaxror:aminrhr? BooleanDtyperrBrr`rr)%rprqr rrrrrrrreturn_distancerrrsrtrvrwrxryrzr{closest_df_idxsna_maskrrrrrrr% df_distance distance_leftdistance_rightr&rrrs% ` r*rrPs|R 1uu))) czz X    {/4})+++%MCc/4})+++%MCc sS#sODDDD sS#sODDDD&  ''+#)   Oaaad#r)GJJ$$.|S$A$ANLLw \ ! $ci10E&F G  \ ! $ci10E&F G  !e 7J#$ IHOOAAAqD$9:HOOAAAqD$9:      g IHOOAAAqD$9:HOOAAAqD$9:      %{2 \ ,!#, t!L!L!x k4HH     && " 1 1!#!}    !e 7K % HOOAAAqD1 2#hooaaad34 5   HOOAAAqD1 2#hooaaad34 5   729m^%DEEANNNlJ#9QQQ !u GJJKs<((C//<63I3IXoaaad34@@d@KK JJJJz7IJJJ $s<((C//<73J3JXoaaad34@@d@KK JJJJz7IJJJ && 8A;  x{1BBMOO T  !e 7 Y ZZ[Q F Mr+c |tn|\|tn|\}}}|dzdzdzi} fdt|jD} | D]} | | | |dz<|r.d|dz| d|dz<d|dz| d|dz<tjtt j|tt j||z} t| dkrtdt|t|d | D|  |t j |t j id ||d d || t| } | | d | jt j| j} | d d |r| d|dzj}| d|dz}tj|D]2} ||| kxx||| kzcc<3|| d|dz<| d|dzgd | S)a Generate a new set of genomic intervals by subtracting the second set of genomic intervals from the first. Parameters ---------- df1, df2 : pandas.DataFrame Two sets of genomic intervals stored as a DataFrame. return_index : bool Whether to return the indices of the original intervals ('index'+suffixes[0]), and the indices of any sub-intervals split by subtraction ('sub_index'+suffixes[1]). Default False. suffixes : (str,str) Suffixes for returned indices. Only alters output if return_index is True. Default ("","_"). cols1, cols2 : (str, str, str) or None The names of columns containing the chromosome, start and end of the genomic intervals, provided separately for each set. The default values are 'chrom', 'start', 'end'. Returns ------- df_subtracted : pandas.DataFrame Notes ----- Resets index, drops completely subtracted (null) intervals, and casts to pd.Int64Dtype(). Nrrc"g|] }|fv | SrVrV)rWirvrxrws r*rYzsubtract..bs)PPPQqc37O7Oq7O7O7Or+r0complement_indexrz*No chromosomes remain after dropping nullscVi|]&}|tjtjj'SrVriinfor@maxrWr4s r* zsubtract..rs)HHH!RXbh//3HHHr+)view_dfr"r7T)rrrrrrrsrtrrr sub_index)rrrruniquerrOrjrrrr?rrrrr:r`r>minrM)rprqrrrsrtryrzr{ name_updatesextra_columns_1r4 all_chroms df_subtractedinds comp_indsrvrxrws @@@r*rr0s*T05})+++%MCc/4})+++%MCc hqk3S#S#L QPPPPP$s{"3"3PPPO **() Q!_%%O.5 .C Wx{*+.@8A;.N Wx{*+ RYs3x(( ) )**T")CHOO > !    < Mt<<<!& c0B0I(J(J'JKM4666UWx{23:!"4x{"BCHHJJ 4 ? ?A dai Idai$8$<$<$>$> > 3<>>3C3C kHQK/0$6!$D#EtTTT r+c|tn|\}}}|tn|\}} } t||d|||} tjtjt || dddf} |j| } | S)a` Generate a new dataframe of genomic intervals by removing any interval from the first dataframe that overlaps an interval from the second dataframe. Parameters ---------- df1, df2 : pandas.DataFrame Two sets of genomic intervals stored as DataFrames. cols1, cols2 : (str, str, str) or None The names of columns containing the chromosome, start and end of the genomic intervals, provided separately for each dataframe. The default values are 'chrom', 'start', 'end'. on : None or list Additional column names to perform clustering on independently, passed as an argument to df.groupby when considering overlaps and must be present in both dataframes. Examples for additional columns include 'strand'. Returns ------- df_setdiff : pandas.DataFrame Nrrr)rrr setdiff1drrjr)rprqrsrtrurvrwrxryrzr{ df_overlappedinds_non_overlapped df_setdiffs r*rrs605})+++%MCc/4})+++%MCc$ SgU%BM,ryS':':M!!!Q$./s9    rx!!%   r+T view_name_colr"r view_regionreturn_as_bool-column view_region already exists in input df)drop_unassigned df_view_colrSr" cols_viewr6rYrSr7r_view)rrleft_onright_onrr])lowerupper)rrrrr>rcrOr:rmake_viewframerdictziprrr is_catalogedrrrrrr2r?rclip)r r<rYrSreturn_view_columnsr"rZr#r$r% df_columns df_trimmed inferred_viewckvskvekvunassigned_intervals lower_vector upper_vectors r*rrsb-1L&(((dJBBBR %%%bj!!JJM   K--//6688?@@    /8/@)+++iMCc)}Cc? fT#sCoB|<<==f>>     : t L L L NLMM M#   !#'     [M222  #'     !!  "J9Z %<%CDD!!F=?U +b"b\9:2r}BMOODEEEb7l+2Lb7l+2L^((|<(PPJrN^((|<(PPJrN&*%%r+c V|tn|\}}}t||||g|6dt||jD}|tn|\}} } t j|||| | gtt|| | g|||g}t||ddd||} | |d |zd |z|g } | d |z|d |z||d id | }tj ||dd | |d j} t!t||} g}| D]}|j|||k}||||gjd \}}}|| vrR| |d i}|t'j|| |j}|j|}t+j||jt0j||jt0j||f\}}|t'jt1j|jd |||j||||d |i}t'j|}||t'j|d}|S)a Find genomic regions in a viewFrame 'view_df' that are not covered by any interval in the dataFrame 'df'. First assigns intervals in 'df' to region in 'view_df', splitting intervals in 'df' as necessary. Parameters ---------- df : pandas.DataFrame view_df : pandas.Dataframe If none, attempts to infer the view from chroms (i.e. df[cols[0]]). view_name_col : str Name of column in view_df with unique reigon names. Default 'name'. cols : (str, str, str) The names of columns containing the chromosome, start and end of the genomic intervals. The default values are 'chrom', 'start', 'end'. cols_view : (str, str, str) or None The names of columns containing the chromosome, start and end of the genomic intervals in the view. The default values are 'chrom', 'start', 'end'. Returns ------- df_complement : pandas.DataFrame Notes ------ Discards null intervals in input, and df_complement has regular int dtype. NcVi|]&}|tjtjj'SrVr7r:s r*r;zcomplement..s)RRR1bhrx((,RRRr+rRrTr)r_df)rrrrrsrtrrTr=r[r)boundsrrL) rrrcrOr:rrbrrcrdrr>rrerarsortedr2rlrrrcomplement_intervalsr?rr@rrr^r_rr`)r r<rSr"rZr#r$r%rkrlrm new_intervalsrr complements group_keyregion_interval region_chrom region_start region_endcomplement_grouprrcomplement_starts_groupcomplement_ends_groups r*rrmsP-1L&(((dJBBBR %%%RRc"R&--//:P6Q6QRRR/8/@)+++iMCc)}Cc? fT#sCoB|<<==f>>   M" Z"_j2o}= dff OR OR =   B  !#  =))0IGM23344JK#-#- !+gm&< &IJ1@"b"1N1UVW1X. lJ I % %.3355<<& 6 =      r|,<== > > > !),3 6-(  ' RL  & &rx 0 0 RL  & &rx 0 0 *-    # !  W4:1=|LLfl ' % 9 <(899+,,,,)K((44$4??K r+c J|tn|\}}} tj||std|} || ||| gdnw|tn|\} } } t j||| | | gtt| | | g||| g}|:t| dgd rtd d}t| ||||| } nzt| |gd std tj | tj| |||| stdtj||jd}| |||i| |<| |||| gd| |j|j} |r| dd| S)a Sorts a bedframe 'df'. If 'view_df' is not provided, sorts by ``cols`` (e.g. "chrom", "start", "end"). If 'view_df' is provided and 'df_view_col' is not provided, uses :func:`bioframe.ops.assign_view` with ``df_view_col='view_region'`` to assign intervals to the view regions with the largest overlap and then sorts. If 'view_df' and 'df_view_col' are both provided, checks if the latter are cataloged in 'view_name_col', and then sorts. df : pandas.DataFrame Valid bedframe. view_df : pandas.DataFrame | dict-like Valid input to make a viewframe. When it is dict-like :func:'bioframe.make_viewframe' will be used to convert to viewframe. If view_df is not provided df is sorted by chrom and start. reset_index : bool Default True. df_view_col: None | str Column from 'df' used to associate intervals with view regions. The associated region in 'view_df' is then used for sorting. If None, :func:'bioframe.assign_view' will be used to assign view regions. Default None. view_name_col: str Column from view_df with names of regions. Default `name`. cols : (str, str, str) or None The names of columns containing the chromosome, start and end of the genomic intervals. The default values are 'chrom', 'start', 'end'. cols_view : (str, str, str) or None The names of columns containing the chromosome, start and end of the genomic intervals in the view. The default values are 'chrom', 'start', 'end'. Returns ------- out_df : sorted bedframe Notes ------- df_view_col is currently returned as an ordered categorical Nrz!not a valid bedframe, cannot sortT)rrRrrTrUrW)rYrSr"rZz9column 'df_view_col' not in input df, cannot sort by view)rYrSz=intervals in df not cataloged in view_df, cannot sort by view) categoriesorderedr)rrr9rr>rrrbrrcrdrrrerrrCategoricalDtyper:r?rr;r`)r r<r`rYrSr"rZrvrwrxrrkrlrmview_cats r*rrsv04|)+++MCc  bt , , ,><=== WWYYFCc?D99994=3D-///) S#- =S#   &c3S/Cc?CCDD& E E   v tLLL R !PQQQ'K '+ FF#6K=NNN  O&{ 3 < < > >???@'+   !S&}-4d   %[188+x9PQQ{Kc37FFFBJ  & &ry 1 1F44d333 Mr+rTc |tn|\}}} |}|ddtj|d|t j|||}t||ddddd|| } | d | z| d |zz | d <| d d  d dd } | |dz|id|r@| j tj | djdkddf} | ddgt!|j|} | | S)a Associates genomic intervals in bedframe ``df`` with regions in viewframe ``view_df``, based on their largest overlap. Parameters ---------- df : pandas.DataFrame view_df : pandas.DataFrame ViewFrame specifying region start and ends for assignment. Attempts to convert dictionary and pd.Series formats to viewFrames. drop_unassigned : bool If True, drop intervals in df that do not overlap a region in the view. Default False. df_view_col : str The column of ``df`` used to specify view regions. The associated region in view_df is then used for trimming. If no view_df is provided, uses the chrom column, ``df[cols[0]]``. Default "view_region". view_name_col : str Column of ``view_df`` with region names. Default 'name'. cols : (str, str, str) or None The names of columns containing the chromosome, start and end of the genomic intervals. The default values are 'chrom', 'start', 'end'. cols_view : (str, str, str) or None The names of columns containing the chromosome, start and end of the genomic intervals in the view. The default values are 'chrom', 'start', 'end'. Returns ------- out_df : dataframe with an associated view region for each interval in ``out_df[view_name_col]``. Notes ------- Resets index. NTrr5rRr7r\F)rrrrrrrsrtroverlap_length) ascendingr0first)keepr]r=rrr)rr>r`rr9rrbrrdrop_duplicatesrrrrrr:rr) r r<rXrYrSr"rZrvrwrx overlap_viewr return_colss r*rrdsl04|)+++MCc BNN4dN+++ r48888)}9G     L Z#%&j36F)GG!"   !1U CC w / / W    MM=72K@$MOOOIRWV__00a088?1DaaaGH t$///2D$$2k2K + r+rU)NNr3N)r7NNN) r7TFFrNNNNT)rNNTTT)rNN)rTNN) NrFFFNNNN)NrFFFNNTFTFrNN)FrNN)NNN)rTNNN)NNrOFNN)NrONN)NTNrONN)FrTrONN)4numpyrpandasrcorerrr core.specsrrcore.stringopsr __all__r r r r rrr_int8 Int8Dtypeint16 Int16Dtypeint32 Int32Dtyper@ruint8 UInt8Dtypeuint16 UInt16Dtypeuint32 UInt32Dtypeuint64 UInt64Dtyperrrrrrr#rrrrrrrrrVr+r*rs ..........>>>>>>>>((((((   *%%%%P8888.333301111DKKKK\PPPPh BHRW|r|~~ BHRX   BHRX   BHRX   BHRX   BHRY)) BHRY)) BHRY)) 000     EEEET  !UUUUp@@@@L   JJJJ^   RRRRn    ]]]]F    WWWWt####R   CCCCP  t&t&t&t&nH  rrrrp ^^^^^^r+