ޚj6ddlmZddlmZmZddlmZddlmZddl m Z ddl m Z dd l mZer dd lmZdd lmZd d gZdd dedede fdZdede fdede ffdZGdd ZGdd Zy))Callable) TYPE_CHECKINGoverload)ContributionsAcceptedError)Column) DataFrame)_to_column_expr) NumericType) ColumnOrName) SparkSession GroupedDataGroupingselfnamecolsreturnc.djt|}|jrt|jnd}|jj }|j j j||||}t||jS)N,) function_name function_aggr group_exprprojected_columns) joinlist _groupingstr get_columns_dfrelationapplyr session)rrr expressionsgroup_by projectionsjdfs fC:\Users\Luka\AppData\Local\Temp\pip-target-o84fipzb\lib\python\duckdb/experimental/spark/sql/group.py _api_internalr*#sz((4:&K&*nns4>>""H..,,.K ((   ! !!% " C S$,, ''f.ctdddtdtffd }j|_j|_|S)Nrrrrc6j}t||g|SN)__name__r*)rrrr,s r)_apizdf_varargs_api.._api1szzT4/$//r+)rr r0__doc__)r,r1s` r)df_varargs_apir30s:0=000JJDM99DL Kr+c,eZdZddZdefdZdefdZy)rrNcd|_|Dcgc] }t|c}|_d|vr|d}ddg}||vsJ||_yycc}w)Nrspecialcuberollup)_typer _cols)rrkwargsxr6accepted_specials r)__init__zGrouping.__init__;sa 267$Qoa($7  Y'G &1 .. .. DJ 8sActdj|jDcgc] }t|c}}|Scc}w)Nr)rr:r)rr<columnss r)r zGrouping.get_columnsDs1((DJJ7JqCFJ788s5ch|j}|jr|jdz|zdzS|S)N())r r9)rr@s r)__str__zGrouping.__str__Hs5""$ ::::#g-3 3r+)rr rN)r0 __module__ __qualname__r>rr rDr+r)rr:s!Sr+ceZdZdZdededdfdZdefdZdefdZ e d edefd Z d edefd Z e d edefd Z e d edefd Ze d edefdZededefdZedeeefdefdZdeeeefzdefdZy)rzjA set of methods for aggregations on a :class:`DataFrame`, created by :func:`DataFrame.groupBy`. groupingdfrNcB||_||_|j|_yr/)rr!r$)rrIrJs r)r>zGroupedData.__init__Us!%'ZZ r+c,t|jSr/)rr!rs r)__repr__zGroupedData.__repr__Zs488}r+c:t|djddS)aCounts the number of records for each group. Examples: -------- >>> df = spark.createDataFrame( ... [(2, "Alice"), (3, "Alice"), (5, "Bob"), (10, "Bob")], ["age", "name"] ... ) >>> df.show() +---+-----+ |age| name| +---+-----+ | 2|Alice| | 3|Alice| | 5| Bob| | 10| Bob| +---+-----+ Group-by name, and count each group. >>> df.groupBy(df.name).count().sort("name").show() +-----+-----+ | name|count| +-----+-----+ |Alice| 2| | Bob| 2| +-----+-----+ countz count_star())r*withColumnRenamedrMs r)rPzGroupedData.count]s8T7+==ngVVr+rcy)zComputes average values for each numeric columns for each group. :func:`mean` is an alias for :func:`avg`. Parameters ---------- cols : str column names. Non-numeric columns are ignored. NrGrrs r)meanzGroupedData.mean{r+ct|}t|dk(rT|jj}|jDcgc])}t |j ts|j+}}t|dg|Scc}w)aComputes average values for each numeric columns for each group. :func:`mean` is an alias for :func:`avg`. Parameters ---------- cols : str column names. Non-numeric columns are ignored. Examples: -------- >>> df = spark.createDataFrame( ... [(2, "Alice", 80), (3, "Alice", 100), (5, "Bob", 120), (10, "Bob", 140)], ... ["age", "name", "height"], ... ) >>> df.show() +---+-----+------+ |age| name|height| +---+-----+------+ | 2|Alice| 80| | 3|Alice| 100| | 5| Bob| 120| | 10| Bob| 140| +---+-----+------+ Group-by name, and calculate the mean of the age in each group. >>> df.groupBy("name").avg("age").sort("name").show() +-----+--------+ | name|avg(age)| +-----+--------+ |Alice| 2.5| | Bob| 7.5| +-----+--------+ Calculate the mean of the age and height in all data. >>> df.groupBy().avg("age", "height").show() +--------+-----------+ |avg(age)|avg(height)| +--------+-----------+ | 5.0| 110.0| +--------+-----------+ ravg) rlenr!schemafields isinstancedataTyper rr*)rrr@rYr<s r)rWzGroupedData.avgsiZt* w<1 XX__F28--!g-Q:ajjZeCf!&&-G!gT53733"hs A;A;cy)aAComputes the max value for each numeric columns for each group. Examples: -------- >>> df = spark.createDataFrame( ... [(2, "Alice", 80), (3, "Alice", 100), (5, "Bob", 120), (10, "Bob", 140)], ... ["age", "name", "height"], ... ) >>> df.show() +---+-----+------+ |age| name|height| +---+-----+------+ | 2|Alice| 80| | 3|Alice| 100| | 5| Bob| 120| | 10| Bob| 140| +---+-----+------+ Group-by name, and calculate the max of the age in each group. >>> df.groupBy("name").max("age").sort("name").show() +-----+--------+ | name|max(age)| +-----+--------+ |Alice| 3| | Bob| 10| +-----+--------+ Calculate the max of the age and height in all data. >>> df.groupBy().max("age", "height").show() +--------+-----------+ |max(age)|max(height)| +--------+-----------+ | 10| 140| +--------+-----------+ NrGrSs r)maxzGroupedData.maxrUr+cy)aComputes the min value for each numeric column for each group. Parameters ---------- cols : str column names. Non-numeric columns are ignored. Examples: -------- >>> df = spark.createDataFrame( ... [(2, "Alice", 80), (3, "Alice", 100), (5, "Bob", 120), (10, "Bob", 140)], ... ["age", "name", "height"], ... ) >>> df.show() +---+-----+------+ |age| name|height| +---+-----+------+ | 2|Alice| 80| | 3|Alice| 100| | 5| Bob| 120| | 10| Bob| 140| +---+-----+------+ Group-by name, and calculate the min of the age in each group. >>> df.groupBy("name").min("age").sort("name").show() +-----+--------+ | name|min(age)| +-----+--------+ |Alice| 2| | Bob| 5| +-----+--------+ Calculate the min of the age and height in all data. >>> df.groupBy().min("age", "height").show() +--------+-----------+ |min(age)|min(height)| +--------+-----------+ | 2| 80| +--------+-----------+ NrGrSs r)minzGroupedData.minrUr+cy)aComputes the sum for each numeric columns for each group. Parameters ---------- cols : str column names. Non-numeric columns are ignored. Examples: -------- >>> df = spark.createDataFrame( ... [(2, "Alice", 80), (3, "Alice", 100), (5, "Bob", 120), (10, "Bob", 140)], ... ["age", "name", "height"], ... ) >>> df.show() +---+-----+------+ |age| name|height| +---+-----+------+ | 2|Alice| 80| | 3|Alice| 100| | 5| Bob| 120| | 10| Bob| 140| +---+-----+------+ Group-by name, and calculate the sum of the age in each group. >>> df.groupBy("name").sum("age").sort("name").show() +-----+--------+ | name|sum(age)| +-----+--------+ |Alice| 5| | Bob| 15| +-----+--------+ Calculate the sum of the age and height in all data. >>> df.groupBy().sum("age", "height").show() +--------+-----------+ |sum(age)|sum(height)| +--------+-----------+ | 20| 440| +--------+-----------+ NrGrSs r)sumzGroupedData.sumrUr+exprscyr/rG)rrcs r)aggzGroupedData.agg=s03r+_GroupedData__exprscyr/rG)rrfs r)rezGroupedData.agg@s9>> from pyspark.sql import functions as F >>> from pyspark.sql.functions import pandas_udf, PandasUDFType >>> df = spark.createDataFrame( ... [(2, "Alice"), (3, "Alice"), (5, "Bob"), (10, "Bob")], ["age", "name"] ... ) >>> df.show() +---+-----+ |age| name| +---+-----+ | 2|Alice| | 3|Alice| | 5| Bob| | 10| Bob| +---+-----+ Group-by name, and count each group. >>> df.groupBy(df.name) GroupedData[grouping...: [name...], value: [age: bigint, name: string], type: GroupBy] >>> df.groupBy(df.name).agg({"*": "count"}).sort("name").show() +-----+--------+ | name|count(1)| +-----+--------+ |Alice| 2| | Bob| 2| +-----+--------+ Group-by name, and calculate the minimum age. >>> df.groupBy(df.name).agg(F.min(df.age)).sort("name").show() +-----+--------+ | name|min(age)| +-----+--------+ |Alice| 2| | Bob| 5| +-----+--------+ Same as above but uses pandas UDF. >>> @pandas_udf("int", PandasUDFType.GROUPED_AGG) # doctest: +SKIP ... def min_udf(v): ... return v.min() >>> df.groupBy(df.name).agg(min_udf(df.age)).sort("name").show() # doctest: +SKIP +-----+------------+ | name|min_udf(age)| +-----+------------+ |Alice| 2| | Bob| 5| +-----+------------+ zexprs should not be emptyrrc3<K|]}t|tywr/)r[r ).0cs r) z"GroupedData.agg..sZ Z<t~~334K   616 74>>*H*$((##**KIICdll++ 7s8C#)r0rErFr2rr r>rrNrPr3rTrWr^r`rbrr rernrGr+r)rrOs7 00y0T0 #WyW<  #  )    242424h% % % % N* * * * X* * * * X3&3Y33 <4S>1d,id,r+N)collections.abcrtypingrr exceptionrcolumnr dataframer functionsr typesr _typingr r$r__all__rr*r3rrrGr+r)r}s$%*2 &%% * % ( (S ( ( (hsI~.8CN3K*X,X,r+