ctjR?<UddlmZddlZddlmZddlmZddlmZddl m Z m Z m Z m Z ddlmZddlmZdd lmZdd lmZdd lmZe r?dd lmZmZmZmZmZdd l mZddlZ ddl!m"Z#ddl$m%Z%ddl&m'Z'm(Z(ddl)m*Z*ddl+m,Z,dZ-de.d<dZ/de.d<e dZ0de.d<e ddddddd d!d"d#d$d%d&d'd(d)d*d+d,e0fZ1de.d-< d.Z2de.d/< e Z3de.d0< dd1dd2Z4d3e.d4<ed56dFd;Z5Gd<d=Z6Gd>d?ed@dAe1fZ7dGdCZ8dHdEZ9dS)I) annotationsN) lru_cache)chain) methodcaller) TYPE_CHECKINGAnyClassVarLiteral) EagerGroupBy) issue_warning)!evaluate_output_names_and_aliases)make_group_by_kwargs)is_pandas_like_dataframe)CallableIterableIteratorMappingSequence) TypeAlias)DataFrameGroupBy)Unpack)NarwhalsAggregation ScalarKwargs)PandasLikeDataFrame)PandasLikeExprz._NativeGroupBy[tuple[str, ...], Literal[True]]r NativeGroupByz(Callable[[pd.DataFrame], pd.Series[Any]] NativeApply)covskewInefficientNativeAggregationanyallcountidxmaxidxminmaxmeanmedianminmodenthnuniqueprodquantilesemsizestdsumvarNativeAggregationz.Callable[[Any], pd.DataFrame | pd.Series[Any]] _NativeAggNonStrHashable)firstlast any_valuez,Mapping[NarwhalsAggregation, Literal[0, -1]]_REMAP_ORDERED_INDEX )maxsizenamekwdsUnpack[ScalarKwargs]returnc |dkrt|dS|dkr*d|vsJd|vsJt||d|dS|r|ddkrt|St|fi|S) Nr,F)dropnar. interpolation)qrDddof)rget)r>r?s \/var/www/html/CA-Chatbot/venv/lib/python3.11/site-packages/narwhals/_pandas_like/group_by.py _native_aggrJEs yD//// zT!!!!$&&&&DD$4DDYZZZZ "488F##q((D!!!  % % % %%ceZdZUdZded<ded<ded<dd Zdd ZddZddZddZ ddZ ddZ ddZ ddZ eddZd dZdS)!AggExpraMWrapper storing the intermediate state per-`PandasLikeExpr`. There's a lot of edge cases to handle, so aim to evaluate as little as possible - and store anything that's needed twice. Warning: While a `PandasLikeExpr` can be reused - this wrapper is valid **only** in a single `.agg(...)` operation. rexprz Sequence[str] output_namesaliasesrANonec>||_d|_d|_d|_dS)N)rNrOrP _leaf_name)selfrNs rI__init__zAggExpr.__init__as#  57rKgroup_byPandasLikeGroupBych|j}|j}t|j||\|_|_|S)zd**Mutating operation**. Stores the results of `evaluate_output_names_and_aliases`. ) compliantexcluder rNrOrP)rVrXdfr\s rIwith_expand_nameszAggExpr.with_expand_namesgs=  "*K Ir7+ + '4< rKpd.DataFrame | pd.Series[Any]c |j}|j}|r*|r|}nW|rQ||jfd|D}n|r|j}| |j }| dx}dkrd|d|j d}t|t|} |j|j|jc | fd| D}n'|s(|s|r||g|j|}|jj } | } | r#| dkr||jd nh||j}nMt3|d kr|d nt|} ||| }t5|rt|j|_n|jd |_|S) z8Evaluate the wrapped expression as a group_by operation.cjg|]/}|j0SrS) from_nativealiasnative).0r>ns result_singles rI z)AggExpr._getitem_aggs..~s6TTTd ..44T::ATTTrKkeepr!z`Expr.mode(keep='z7')` is not implemented in group by context for backend z3 Hint: Use `nw.col(...).mode(keep='any')` instead.cg|]}jg|fid|jfi|dS)F) ascendingrG)groupbyr0 sort_values reset_indexhead sort_index)recolkeyskwargsrds rIrhz)AggExpr._getitem_aggs..s   NFN)rVrXgroupednamesresultr[ node_kwargsrimsgcolsimplbackend_versionselectrrrsrdrfrgs @@@@@rI _getitem_aggszAggExpr._getitem_aggsss#! ;;==1 8T77991 8\\^^FF [[]]/ 8#LLNNM#::<8+DLD&1133B**       $     FF\\^^ 8t}} 8$2C2C2E2E 8&T__&&w/H/H%/H'IJJF%5D"3355O~~ :Of$<$<   >>>>))(.99!$UqU1XXd5kkF&T__&&wv77F #F + + *!$,//FNN,q/FK rKboolc|jdkS)Nr leaf_namerVs rIryzAggExpr.is_lens~&&rKc|jdkS)Nr9rrs rIrzAggExpr.is_last~''rKc|jdkS)Nr8rrs rIrzAggExpr.is_firsts~((rKc|jdkS)Nr*rrs rIr}zAggExpr.is_moderrKc|jdkS)Nr:rrs rIrzAggExpr.is_any_values~,,rKcztt|jjdkS)NrG)rrrN _metadataop_nodes_reversedrs rIrzzAggExpr.is_top_level_functions.4 +==??@@AAQFFrKNarwhalsAggregation | Anycn|jx}r|St|j|_|jSN)rUrYrN)rVr>s rIrzAggExpr.leaf_names5? "4 K+66tyAArKr5cnt|j}t|jj}|jtvrL|j drd}t|tdt|jSt|fi|jS)z@Return a partial `DataFrameGroupBy` method, missing only `self`. ignore_nullszd`Expr.any_value(ignore_nulls=True)` is not supported in a `group_by` context for pandas-like backendr+)n) rY_remap_expr_namernextrNrrr;rsrHrrrJ)rV native_name last_noders rIrzAggExpr.native_aggs'88HH ,>>@@AA >1 1 1##N33 /6*#...)=dn)MNNN N;;;)*:;;;rKN)rNrrArQ)rXrYrArM)rXrYrAr_)rAr)rAr)rAr5)__name__ __module__ __qualname____doc____annotations__rWr^rryrrr}rrzpropertyrrrSrKrIrMrMRs38888    ;;;;z''''(((())))((((----GGGGX < < < < < d.Z d?d1Z d@d5Z dAd7Z dBd9Z d:S)CrYr2r'r(r&r)r*r1r3rr0n_uniquer,r#r.r"r!r8r+r9r:z9ClassVar[Mapping[NarwhalsAggregation, NativeAggregation]] _REMAP_AGGStuple[str, ...]_original_columnsz list[str]r_output_key_nameszMapping[str, bool]rrAc|jS)z>Group keys to ignore when expanding multi-output aggregations.)_excluders rIr\zPandasLikeGroupBy.excludes }rKr]rrr(Sequence[PandasLikeExpr] | Sequence[str]drop_null_keysrrQct|j|_||_|||\|_|_|_g|j|jR|_t||_ |j j |_ t|j jj|j jr"|j d|_ dSdS)N)rT)drop)tuplerr_drop_null_keys _parse_keys_compliant_framerrrrrr[rd_nativesetindexr intersectionrn)rVr]rrrs rIrWzPandasLikeGroupBy.__init__s"'rz!2!2-DHDTDT E E Atz4+A*P4:)O8N)O)O 4N S S S~, t|!' ( ( 5 5dn6L M M ?<333>>DLLL ? ?rKexprsrcd}g}d}|D]}|t||||sd}t |j}|jddx}r"|r||krd|d|d}t||}|rR|j t|d j |jfi|j} n/|j j |jfi|j} | |_|r|rB|j} | ||} n{|jt| j|j } n5|jjjrt7|| |} |jj} | } | r| d kr| d n| } |!| |S) NTrSForder_byz?Only one `order_by` can be specified in `group_by`. Found both z and .r8) na_position)rrtrv)"appendrMr^ _is_simplerrrrsrHrrrmrrlrcopyrrxr[r{r|r__native_namespace__ DataFramegroupsrdemptyempty_results_error _apply_aggsrrrrn_select_results)rVrall_aggs_are_simple agg_exprsrrNmd_current_order_byrrrfrrrs rIaggzPandasLikeGroupBy.aggs"#%  - -D   WT]]<&&djoo''&B&B+/+@&B&BGG+dl*4:??+<+<VV@UVVG  6 ^::<<..t/A/A)/L/LMM<<>>HH(($*I^ " ( 6%'' '%%gu55F~-//11 >>   *& 8 8   t  , , , ,''))F##FI666rKrSequence[AggExpr] pd.DataFramec tjd|D}|j|djg|j|Rtt|j|j dS)zgResponsible for remapping temp column names back to original. See `ParseKeysGroupBy`. c3$K|] }|jV dSr)rP)rees rI z4PandasLikeGroupBy._select_results..Ls$'E'Ea 'E'E'E'E'E'ErKF)validate_column_names)strict) r from_iterabler[ _with_native simple_selectrrenamedictzipr)rVr]r new_namess rIrz!PandasLikeGroupBy._select_resultsEs''E'E9'E'E'EEE  DN ' '% ' H H  4 J 4)2 4 4 4 VDTZ)?NNNOO P P rKIterable[AggExpr]#list[pd.DataFrame | pd.Series[Any]]c fd|DS)Nc:g|]}|SrS)r)rerrVs rIrhz3PandasLikeGroupBy._getitem_aggs..Vs%555!%%555rKrS)rVrs` rIrzPandasLikeGroupBy._getitem_aggsSs6555u5555rKrrIterable[PandasLikeExpr]ct|jj}||}|j}|r%|dkr ||dS||S)a"Stub issue for `include_groups` [pandas-dev/pandas-stubs#1270]. - [User guide] mentions `include_groups` 4 times without deprecation. - [`DataFrameGroupBy.apply`] doc says the default value of `True` is deprecated since `2.2.0`. - `False` is explicitly the only *non-deprecated* option, but entirely omitted since [pandas-dev/pandas-stubs#1268]. [pandas-dev/pandas-stubs#1270]: https://github.com/pandas-dev/pandas-stubs/issues/1270 [User guide]: https://pandas.pydata.org/pandas-docs/stable/user_guide/groupby.html [`DataFrameGroupBy.apply`]: https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.core.groupby.DataFrameGroupBy.apply.html [pandas-dev/pandas-stubs#1268]: https://github.com/pandas-dev/pandas-stubs/pull/1268 )rF)include_groups)warn_complex_group_byr[r_apply_exprs_functionapplyrr)rVrrrfuncrs rIrzPandasLikeGroupBy._apply_aggsXs ~-))%00  >>   5 5 5 7 76 A A5e444 4uT{{rKrcjjjjdfd }|S)Nr]rrApd.Series[Any]cj|fdD}|r t|ddinggf\}}||jS)Nc3fK|]+}|D]}|jjd|jfV,dS)rN)rdilocr>)rerNrrr[s rIrzFPandasLikeGroupBy._apply_exprs_function..fn..tsg DOO!!$di0rKrT)rcontext)r[rrrd) r]results out_group out_namesr[r into_seriesrfrVs @rIfnz3PandasLikeGroupBy._apply_exprs_function..fnrs33B77I!G BI#V3#=#=#=#=rSUh Iy;y 2FFFM MrK)r]rrAr)r[r{_seriesr)rVrrrrfs`` @@rIrz'PandasLikeGroupBy._apply_exprs_functionns` ^ 2 2 4 4j.  N N N N N N N N N rK)Iterator[tuple[Any, PandasLikeDataFrame]]c#ZK|jj|jfi|j}t j5t jddt|j j }|D]!\}}|||j |j fV" ddddS#1swxYwYdS)Nignorez#.*a length 1 tuple will be returned)messagecategory) rrlrrrwarningscatch_warningsfilterwarnings FutureWarningr[rrr)rVr with_nativekeygroups rI__iter__zPandasLikeGroupBy.__iter__~s.&$,&tz'8'8RRDTUVVVVV W W W W W W W W W W W W W W W W W W WsA B  B$'B$N)rAr)r]rrrrrrrArQ)rrrAr)rrr]rrAr)rrrAr)rrrrrAr)rrrAr)rAr)rrrrrrr\rWrrrrrrrSrKrIrYrYsN uNN (N u N u N  N uN uN vN IN N JN uN uN N !N" U#NK&'&&&EO    8((((K X????,-7-7-7-7^     6666 , W W W W W WrKrYrr ValueErrorc$d}t|S)zJDon't even attempt this, it's way too inconsistent across pandas versions.auNo results for group-by aggregation. Hint: you were probably trying to apply a non-elementary aggregation with a pandas-like API. Please rewrite your query such that group-by aggregations are elementary. For example, instead of: df.group_by('a').agg(nw.col('b').round(2).mean()) use: df.with_columns(nw.col('b').round(2)).group_by('a').agg(nw.col('b').mean()) )r)rs rIrrs ^ c??rKrQc0tdtdS)Na)Found complex group-by expression, which can't be expressed efficiently with the pandas API. If you can, please rewrite your query such that group-by aggregations are simple (e.g. mean, std, min, max, ...). Please see: https://narwhals-dev.github.io/narwhals/concepts/improve_group_by_operation/)r UserWarningrSrKrIrrs) W  rK)r>r4r?r@rAr5)rAr)rArQ): __future__rr functoolsr itertoolsroperatorrtypingrrr r narwhals._compliantr narwhals._exceptionsr narwhals._expression_parsingr narwhals._pandas_like.utilsrnarwhals.dependenciesrcollections.abcrrrrrrpandaspdpandas.api.typingr_NativeGroupBytyping_extensionsrnarwhals._compliant.typingrrnarwhals._pandas_like.dataframernarwhals._pandas_like.exprrrrrr r4r5r6r;rJrMrYrrrSrKrIr's["""""""!!!!!!888888888888,,,,,,......JJJJJJ<<<<<<:::::: POOOOOOOOOOOOOO      DDDDDD((((((LLLLLLLLCCCCCC999999OMOOOOC CCCC*1-*@@@@@&                   '" ,hH HHHH@ 6 FF 2 & & & &D<D<D<D<D<D<D<D