Ë
|
nñúh¹. ã ó d Z ddlmZ ddlmZmZ ddlmZmZm Z ddl
|
Z ddl m Z ddlmZmZmZ ddlmZ dd lmZmZ dd
|
lmZmZmZ dd lm Z dd l m Z dd l m Z erddl!m"Z"m#Z# ddl$m%Z%m&Z& d dZ' G d de« Z( G d de(« Z) G d de(« Z*d dZ+d dZ, d!dZ- d!dZ.d"dZ/ d#d Z0 d$d Z1y)%z
|
Module responsible for execution of NDFrame.describe() method.
|
|
Method NDFrame.describe() delegates actual execution to function describe_ndframe().
|
é )Ú annotations)ÚABCÚabstractmethod)Ú TYPE_CHECKINGÚCallableÚcastN)Ú Timestamp)ÚDtypeObjÚNDFrameTÚnpt)Úvalidate_percentile)Ú is_bool_dtypeÚis_numeric_dtype)Ú
|
ArrowDtypeÚDatetimeTZDtypeÚExtensionDtype)Ú Float64Dtype)Úconcat)Úformat_percentiles)ÚHashableÚSequence)Ú DataFrameÚSeriesc óØ t |« }| j dk( rt t d| « ¬« }nt t d| « ||¬« }|j |¬« }t t |« S )a Describe series or dataframe.
|
|
Called from pandas.core.generic.NDFrame.describe()
|
|
Parameters
|
----------
|
obj: DataFrame or Series
|
Either dataframe or series to be described.
|
include : 'all', list-like of dtypes or None (default), optional
|
A white list of data types to include in the result. Ignored for ``Series``.
|
exclude : list-like of dtypes or None (default), optional,
|
A black list of data types to omit from the result. Ignored for ``Series``.
|
percentiles : list-like of numbers, optional
|
The percentiles to include in the output. All should fall between 0 and 1.
|
The default is ``[.25, .5, .75]``, which returns the 25th, 50th, and
|
75th percentiles.
|
|
Returns
|
-------
|
Dataframe or series description.
|
é r ©Úobjr )r ÚincludeÚexclude)Ú percentiles)Ú_refine_percentilesÚndimÚSeriesDescriberr ÚDataFrameDescriberÚdescriber )r r r r Ú describerÚresults úOH:\Change_password\venv_build\Lib\site-packages\pandas/core/methods/describe.pyÚdescribe_ndframer) 7 sm ô8 & kÓ2Kð xx1}Ü#ÜXsÓ#ô
|
ô 'Ü[ #Ó&ØØô
|
ð × Ñ ¨KÐ Ó 8FÜ &Ó !Ð!ó c ó* e Zd ZdZddZedd« Zy)ÚNDFrameDescriberAbstractzAbstract class for describing dataframe or series.
|
|
Parameters
|
----------
|
obj : Series or DataFrame
|
Object to be described.
|
c ó || _ y ©Nr )Úselfr s r( Ú__init__z!NDFrameDescriberAbstract.__init__n s Ør* c ó y)z¸Do describe either series or dataframe.
|
|
Parameters
|
----------
|
percentiles : list-like of numbers
|
The percentiles to include in the output.
|
N© )r/ r s r( r% z!NDFrameDescriberAbstract.describeq s r* N)r úDataFrame | SeriesÚreturnÚNone)r ú Sequence[float] | np.ndarrayr4 r3 )Ú__name__Ú
|
__module__Ú __qualname__Ú__doc__r0 r r% r2 r* r( r, r, e s ñóð ò ó ñ r* r, c ó$ e Zd ZU dZded< ddZy)r# z2Class responsible for creating series description.r r c óR t | j « } || j |« S r. )Úselect_describe_funcr )r/ r Ú describe_funcs r( r% zSeriesDescriber.describe s& Ü,Ø HHó
|
ñ TXX {Ó3Ð3r* N)r r6 r4 r )r7 r8 r9 r: Ú__annotations__r% r2 r* r( r# r# | s
Ù<à Kô4r* r# c óP e Zd ZU dZded< d fd ZddZd dZ xZS )
|
r$ ab Class responsible for creating dataobj description.
|
|
Parameters
|
----------
|
obj : DataFrame
|
DataFrame to be described.
|
include : 'all', list-like of dtypes or None
|
A white list of data types to include in the result.
|
exclude : list-like of dtypes or None
|
A black list of data types to omit from the result.
|
r r c ó¦ || _ || _ |j dk( r$|j j dk( r t d« t | |« y )Né r z+Cannot describe a DataFrame without columns)r r r" ÚcolumnsÚsizeÚ
|
ValueErrorÚsuperr0 )r/ r r r Ú __class__s r( r0 zDataFrameDescriber.__init__ sH ø ð Ø à 88q=S[[× -Ñ -°Ò 2Ü Ð JÓKÐ Kä ÑÕ r* c óX | j « }g }|j « D ]( \ }}t |« }|j |||« « * t |« }t |D cg c] }|j |d¬« c}dd¬« } |j j « | _ | S c c}w )NF)Úcopyr )ÚaxisÚsort) Ú _select_dataÚitemsr= ÚappendÚreorder_columnsr ÚreindexrC rI )
|
r/ r ÚdataÚldescÚ_Úseriesr> Ú col_namesÚxÚds
|
r( r% zDataFrameDescriber.describe¦ s¢ Ø× Ñ Ó"à Ø ò =IAvÜ 0°Ó 8MØ LL v¨{Ó;Õ <ð =ô $ EÓ* Ü Ø7<Ö =°!QYYy uYÕ -Ò =ØØô
|
ð
|
LL×%Ñ%Ó' Øùò >s Á"B'c ó¬ | j `| j Tt j dg}| j j |¬« }t |j « dk( r | j }|S | j dk( r'| j d}t |« | j }|S | j j | j | j ¬« }|S )z Select columns to be described.Údatetime)r r Úallz*exclude must be None when include is 'all')r r ) r r ÚnpÚnumberr Ú select_dtypesÚlenrC rE )r/ Údefault_includerQ Úmsgs r( rL z DataFrameDescriber._select_data· s à LLÐ t§|¡|Ð';ä46·I±I¸zÐ3JOØ88×)Ñ)°/Ð)ÓBDÜ4<<Ó AÒ%Øxxð ð \\UÒ "Ø||Ð'ØBÜ oÐ%Ø88Dð ð 88×)Ñ)Ø Ø ð *ó Dð r* )r r r ústr | Sequence[str] | Noner ra r4 r5 )r r6 r4 r )r4 r ) r7 r8 r9 r: r? r0 r% rL Ú __classcell__)rG s @r( r$ r$ sD ø
ñ
|
ð
|
Nð à ð ð ,ð ð
|
,ð ð
|
õ ó ÷"r* r$ c ó¸ g }t « }t d | D « t ¬« }|D ]0 }|D ]) }||vs|j |« |j |« + 2 |S )z,Set a convenient order for rows for display.c 3 ó4 K | ] }|j y wr. )Úindex)Ú.0rV s r( ú <genexpr>z"reorder_columns.<locals>.<genexpr>Ð s è ø Ò3¨AGGÑ3ùs )Úkey)ÚsetÚsortedr^ ÚaddrN )rR ÚnamesÚ
|
seen_namesÚ ldesc_indexesÚidxnamesÚnames r( rO rO Ì se à EÜ #£JÜÑ3¨UÔ3¼Ô=MØ!ò #Ø ò #DØ:Ò%ØtÔ$Ø TÕ"ñ #ð#ð
|
Lr* c ó ddl m} t |« }g d¢|z dgz }| j « | j « | j « | j « g| j |« j « z | j « gz }t | j t « r_t | j t « r:| j j dk( rd}nYddl}t |j! « « }n;t# « }n0| j j dv rt% j d« }nd} |||| j& |¬ « S )
|
zÛDescribe series containing numerical data.
|
|
Parameters
|
----------
|
series : Series
|
Series to be described.
|
percentiles : list-like of numbers
|
The percentiles to include in the output.
|
r ©r )ÚcountÚmeanÚstdÚminÚmaxÚmNÚiufbÚfloat©re rp Údtype)Úpandasr r rs rt ru rv ÚquantileÚtolistrw Ú
|
isinstancer| r r ÚkindÚpyarrowÚfloat64r r[ rp )rT r r Úformatted_percentilesÚ
|
stat_indexrW r| Úpas r( Údescribe_numeric_1dr Ù s õ ä .¨{Ó ;Ðâ0Ð3HÑHÈEÈ7ÑRJà ¨¯
|
©
|
« °f·j±j³lÐCØ
|
//+Ó
|
&×
|
-Ñ
|
-Ó
|
/ñ 0à ::<.ñ ð ô &,,¤Ô/Ü fll¤JÔ /Ø||× Ñ CÒ'à ã$ä" 2§:¡:£<Ó0ä NEØ × Ñ fÑ $äÓ!àÙ !:¨F¯K©K¸uÔ EÐEr* c ó< g d¢}| j « }t ||dk7 « }|dkD r!|j d |j d }}d}n"t j
|
t j
|
}}d}| j « |||g}ddlm} | ||| j |¬« S )zãDescribe series containing categorical data.
|
|
Parameters
|
----------
|
data : Series
|
Series to be described.
|
percentiles_ignored : list-like of numbers
|
Ignored, but in place to unify interface.
|
)rs ÚuniqueÚtopÚfreqr NÚobjectrr r{ )
|
Ú value_countsr^ re Úilocr[ Únanrs r} r rp )
|
rQ Úpercentiles_ignoredrl Ú objcountsÚ count_uniquer r r| r' r s
|
r( Údescribe_categorical_1dr s ò /EØ×!Ñ!Ó#IÜy ¨a¡ Ñ0Ó1LØaÒØ OO AÑ&¨ ¯©°qÑ(9TØô FFBFFTØàjjlL¨#¨tÐ 4Få á & ¨D¯I©I¸UÔ CÐCr* c óÆ ddg}| j « }t ||dk7 « }| j « |g}d}|dkD rÜ|j d |j d }}| j
|
j } | j « j j d« }
|
t |« }|j | |j | « }n|j | « }|g d¢z }|||t |
|
j « | ¬« t |
|
j « | ¬« gz }n,|dd gz }|t j" t j" gz }d
|
}dd lm} | ||| j( |¬ « S ) zøDescribe series containing timestamp data treated as categorical.
|
|
Parameters
|
----------
|
data : Series
|
Series to be described.
|
percentiles_ignored : list-like of numbers
|
Ignored, but in place to unify interface.
|
rs r r NÚi8)r r ÚfirstÚlast)Útzr r r rr r{ )r r^ rs re r Údtr ÚdropnaÚvaluesÚviewr ÚtzinfoÚ
|
tz_convertÚ tz_localizerv rw r[ r r} r rp ) rQ r rl r r r' r| r r r Úasintr s r( Ú$describe_timestamp_as_categorical_1dr¡ " sG ð hÐ EØ×!Ñ!Ó#IÜy ¨a¡ Ñ0Ó1LØ'+§z¡z£|°\Ð&BFØ EØaÒØ OO AÑ&¨ ¯©°qÑ(9TØ WWZZØ ×$Ñ$×)Ñ)¨$Ó/ÜnØ ::Ð ! b nà.. Ó$Cà// "Ó%CØ Ò1Ñ1ØØ Ø Ü eiik bÔ )Ü eiik bÔ )ð
|
ñ
|
ð % Ñ Ø266266Ð"Ñ"Øå á & ¨D¯I©I¸UÔ CÐCr* c ó& ddl m} t |« }g d¢|z dgz }| j « | j « | j « g| j |« j « z | j « gz } |||| j ¬« S )zÛDescribe series containing datetime64 dtype.
|
|
Parameters
|
----------
|
data : Series
|
Series to be described.
|
percentiles : list-like of numbers
|
The percentiles to include in the output.
|
r rr )rs rt rv rw )re rp )
|
r} r r rs rt rv r~ r rw rp )rQ r r r r
rW s r( Údescribe_timestamp_1dr£ R s õ ä .¨{Ó ;Ðâ)Ð,AÑAÀUÀGÑKJà tyy{ D§H¡H£JÐ/Ø
|
-- Ó
|
$×
|
+Ñ
|
+Ó
|
-ñ .à 88:,ñ ð ñ
|
!:¨D¯I©IÔ 6Ð6r* c ó t | j « rt S t | « rt S | j j
|
dk( st | j t « rt S | j j
|
dk( rt S t S )zSelect proper function for describing series based on data type.
|
|
Parameters
|
----------
|
data : Series
|
Series to be described.
|
ÚMrx ) r r| r r r r r r r£ )rQ s r( r= r= j sc ô TZZÔ Ü&Ð&Ü $Ô Ü"Ð"Ø CÒ ¤:¨d¯j©j¼/Ô#JÜ$Ð$Ø CÒ Ü"Ð"ä&Ð&r* c ó. | t j g d¢« S t | « } t | « d| vr| j d« t j
|
| « } t j | « }| J t |« t | « k r t d« |S )zº
|
Ensure that percentiles are unique and sorted.
|
|
Parameters
|
----------
|
percentiles : list-like of numbers, optional
|
The percentiles to include in the output.
|
)g Ð?ç à?g è?r§ z%percentiles cannot contain duplicates) r[ ÚarrayÚlistr rN Úasarrayr r^ rE )r Ú unique_pctss r( r! r! s ð ÐÜxxÒ)Ó*Ð*ô {Ó#Kô Ô$ð +Ñ Ø×Ñ3Ô ä**[Ó)Kô ))KÓ(KØ Ð "Ð"Ð "Ü
|
;Ó#kÓ*Ò*ÜÐ@ÓAÐAà Ðr* )
|
r r r ra r ra r ú#Sequence[float] | np.ndarray | Noner4 r )rR zSequence[Series]r4 zlist[Hashable])rT r r úSequence[float]r4 r )rQ r r r r4 r )rQ r r r r4 r )rQ r r4 r )r r¬ r4 znpt.NDArray[np.float64])2r: Ú
|
__future__r Úabcr r Útypingr r r Únumpyr[ Úpandas._libs.tslibsr Úpandas._typingr
|
r r Úpandas.util._validatorsr Úpandas.core.dtypes.commonr r Úpandas.core.dtypes.dtypesr r r Úpandas.core.arrays.floatingr Úpandas.core.reshape.concatr Úpandas.io.formats.formatr Úcollections.abcr r r} r r r) r, r# r$ rO r r r¡ r£ r= r! r2 r* r( ú<module>r» s> ðñõ
|
#÷÷ñ ó å )÷ñ õ
|
8÷÷ñ õ 5Ý -å 7á÷÷
|
ð +"à ð+"ð (ð+"ð (ð +"ð
|
5ð +"ð ó +"ô\ sô ô. 4Ð.ô 4ôAÐ1ô AóH
|
ó&FðR DØ
|
ð Dà(ð Dð ó Dð@-DØ
|
ð-Dà(ð-Dð ó-Dó`7ð0'Ø
|
ð'à ó'ð, Ø4ð à ô r*
|
