从电子表格到R语言：初学者的R和RStudio指南

R语言

需积分: 9 55 浏览量更新于2024-07-17 收藏 12.54MB PDF 举报

身份认证购VIP最低享 7 折!

30元优惠券

资源详情

资源推荐

programminglanguage.Fornow,let’sgothroughwhatitmeansto“havesomedata.”

1.1 Data: What, where, how?

Isaidyouhavesomedatathatyouwanttodosomethingwith,whichwasn’tavery

precisestatement.Thatwasintentional.Iguaranteeyouhavesomedataevenifyou

don’trealizeit.Youmaybethinkingthatdataisexclusivelywhateverisstoredinyour

Excelfile,butdataismuchmorethanthat.Weallhavedata,becauseit’severywhere.

Beforeyougoanalyzingyourowndata,it’simportanttorecognizeitsstructure(bothas

youunderstandit,andasRwill)sothatyoubeginwithasolidfoundationofwhatit

meanstohavesomedata.

1.1.1 WHAT IS DATA?

Dataexistsinmanyforms,notjustasnumbersandlettersinaspreadsheet.Itmay

alsobestoredinadifferentfiletype,suchascommaseparatedvalues(CSV),as

wordsinabook,orasvaluesinatableonawebpage.

NOTEIt’scommontostorecommaseparatedvaluesina.csvfile.Thisformatis

particularlyusefulbecauseit’splaintext—valuesseparatedbycommas.We’llreturn

towhythat’susefulinsection1.1.6.

Datamaynotbestoredatall—streamingdatacomesasaflowofinformation,such

asthesignalyourTVpicksupandprocesses,yourTwitterfeed,ortheoutputfroma

measuringdevice.Wecanstorethisdataifwewantto,butoftenwewantto

understandtheflowasit’shappening.

Dataisn’talwayspretty(infact,mosttimesit’sdirty,mundane,andseemingly

uninteresting),anditisn’talwaysintheformatwewant.Havingsometoolsonhandto

managedataisapowerfuladvantageandiscriticaltoachievingareliablegoal,but

that’sonlyusefulifyouknowwhatyourdatarepresentsbeforeyoudoanythingfurther

withit.“Garbagein,garbageout”warnsthatyoucan’tperformananalysisonterrible

dataandexpecttogetameaningfulresult.Youmayverywellhavetriedtoevaluatea

calculationinExcelonlytohavetheresultshowupas#VALUE!becauseyoutriedto

divideanumberbysometext,eventhoughthat“text”lookedlikenumbers.Thetypes

ofyourvalues(text,numbers,images,andsoon)arethemselvespiecesofdatawith

possiblemeaningsbehindthem,andyou’lllearnhowtobestmakeuseofthem.

Sowhatis“gooddata”?Whatdothevaluesyouhaverepresent?

1.1.2 SEEING THE WORLD AS DATA SOURCES

Weexperiencetheworldthroughoursenses—touching,seeing,hearing,tasting,

smelling,andgenerallyabsorbinglifearoundus.Eachofthoseinputchannelshandles

availabledata,andourbrainsprocessthem,mixingthesignalstogethertoformour

pictureoftheworldinabrilliantlycomplexwaythatweconstantlytakeforgranted.

Everytimeyouuseanyofyoursenses,you’retakingameasurementoftheworld.How

brightisthesuntoday?Isacarapproaching?Issomethingburning?Isthereenough

coffeeleftinthepotforanothercup?Weconstructmeasuringtoolstomakelifeeasier

forusandhandlesomeofthedataconsistently—thermometerstomeasure

temperatures,scalestomeasureweights,rulerstomeasurelengths.

Wegoastepfurtherandcreatemoretoolstosummarizethatdata—carinstrument

panelstosimplifytheinternalmeasurementsoftheengine;weatherstationsto

summarizetemperature,wind,andpressure.Withthedigitalage,wenowhavean

overloadofdatasourcesatourdisposal.Theinternetprovidesdataonvirtuallyany

andallaspectsoftheworldwemightbeinterestedin,andwecreatemoretoolsto

managethese—weather,finance,socialmedia,thenumberofastronautscurrentlyin

space(www.howmanypeopleareinspacerightnow.com),listsofepisodesofThe

Simpsons,allavailableatourdisposal.Theworldistrulymadeupofdata.

That’snottosaythedataisinanywayfinite.Weconstantlyaddtotheavailable

sourcesofdata,andbyaskingnewquestionswecanidentifynewdatawewantto

obtain.Dataitselfalsogeneratesmoredata.Metadataistheadditionaldatathat

describessomeotherdata—thenumberofsubjectsinatrial,theunitsofa

measurement,thetimeatwhichasamplewastaken,thewebsitefromwhichthedata

wascollected.Allthesearedatatooandneedtobestored,maintained,andupdated

astheychange.

Youinteractwithdatainvariouswaysallthetime.Oneofthegreatestachievementsof

theWorldWideWebhasbeentogather,collate,andsummarizeourdataforusin

moreeasilydigestibleforms.Thinkabouthowyouwouldhaverequestedataxi20

yearsago,beforetheriseofsmartphonesandtheappecosystem.You’dlookupthe

phonenumberofataxicompany,phonethem,tellthedispatcherwhereyouwereor

wouldbe,whereyouwantedtogo,andwhattimeyouwantedtobepickedup.The

dispatcherwouldsendouttherequesttoalldrivers,oneofwhomwouldacceptthe

request.Attheendofyourjourney,you’dpaywithcashoracardtransactionand

receiveareceipt.

Now,withthedigitalconnectionsbetweendevices,continuousinternetaccess,and

GPStracking,thatprocesssimplifiestoopeningarideshareapp,enteringyour

destination,andreceivingafareestimate,becauseyourphonealreadyknowswhere

youare.Therideshareprogramreceivesthisdataandselectsanappropriately

close/availabledriver,exchangesyourcontactdetailsincaseanyoneneedsthem,and

routesthedrivertoyou.Attheendofyourjourney,youraccountischargedthe

appropriateamount,andareceiptisemailedtoyou.

Inbothcases,thesamedataflowedbetweenalltheparties.Inthelatter,fewerpeople

neededtobeinvolvedbecausethecomputersystemshaveaccesstotherelevant

data.Yourphonecommunicateswiththerideshareserver,yourphonecommunicates

withtheGPSsystemtolocateitself,andtherideshareservercommunicateswitha

paymentservertoauthorizepaymentandtheemailservertosendthereceipt.

Ateverypointalongtheway,variousdatacanbecollected(anonymously,where

required)andsavedforlateranalysis.Howmanypeoplerequestedridestotheairport

thismonth?Whatwastheaveragedistancetravelled?Whatwastheaveragewait

time?DopeoplerequestmoreexpensivetripsfromAppleorAndroiddevices?Someof

thiswasavailablepreviously,butithasneverbeeneasiertoaggregateandcompare.

Manybusinessesopenupaccesstothirdpartydevelopersusinganapplication

programminginterface(API)sothatthedatacanbemoresystematicallyaccessed.For

example,UberhasanAPIthatallowssoftwaretoaskforfareestimatesorride

histories(withauthentication,toapprovedaccounts).Thisishowyourphoneappis

abletocommunicatewiththeUberservers.Sureenough,someonehaswrittenanR

packagetoworkwiththisAPI,meaningyoucanincludedatadirectfromUberinyour

analysis,or(intheory)requestaridedirectfromR.

NOTEGoodsoftwarehasadocumentedwaytointeractwithitsothatusersandthe

softwareareabletocommunicateclearlyandeffectively.Thiscandescriberequeststhat

canbesenttoaserver(andtheexpectedresponses)orjusthowafunctionshouldbe

used(andtheexpectedreturnvalue).

1.1.3 DATA MUNGING

Datamungingreferstothecleaningupandpreparationofdata.Mostdatacollected

isn’treadytobeusedinananalysisorpresentation.Usuallythereareinputsto

validate,summariestocalculate,valuestocombineorremove,orrestructuringto

perform.Thisisacommonlyoverlookedaspectofusingdataforscience,butit’sofvital

importance.Failingtoproperlyhandledatacanleadtodifficultyworkingwithitand,

worse,incorrectconclusionsdrawnfromit.

Thetermsdatamunging,datawrangling,datascience,dataanalysis,datahazmat,

andmanyothersareallnamesformoreorlessthesamething,withdifferent

emphasesanddifferenttrajectoriesdependingonwherethedataiscomingfromor

goingto.Mostanalyses(betheyelaborate,sophisticatedregressions,orsimple

visualizations)beginwithsomeformofdatamunging.Oftenthat’smerelyreadingthe

dataintosoftware,inwhichcasesomeofthehandlingisperformedonyourbehalfwith

assumptions(thesevaluesaretreatedasdates,theseaswords,andsoon).Having

thepowertocontrolhowthathandlingisperformedcanbeessentialwhenthose

assumptionsarebroken,orwhenyouwanttotreatyourdatainaparticularway.

Anytimeyouhavegroupsofrecordsinyourdata,whetheryears,patients,animals,

colors,vehicles,oranythingelse,andyouneedtotreatthemdifferently(coloralinea

certainway,onlyincluderecordsinanaverageofsimilarthings,calculatehowa

quantityhaschangedbetweengroups),you’llperformdatamungingbecauseyouneed

toallocaterecordstoaparticulargroupsomehow.Anyothertransformation,cleaning,

orprocessingofthedataalsocountstowarddatamunging.Itquicklybecomes

apparentthatalargeportionofanyanalysiscan(orshould)involvealotofdata

mungingifitsconclusionsaretobetrusted.

1.1.4 WHAT YOU CAN DO WITH WELL-HANDLED DATA

Ihopeit’sclearbythispointthatdataispotentiallyofgreatimportance.Itisroutinely

morethanjustnumbersinatable.Medicaldataoftenrepresentsrealhumanlivesand

theeffectaparticularinterventionhashad,bethatlifesavinglypositiveortragically

negative.Theseeffectsaren’talwaysimmediatelyobvioustosomeoneviewingthem

fromagivenperspective,soit’stheroleofthedataanalyst(professionalorincidental)

toextractpatternsfromdatainordertomakeadecision.

Analysisofdataisoftenusefulinextractingnonobviouspatterns.Forexample,

althoughyoumayrecognizeapatterntothesequence

#>2468101214161820

(countingbytwos),itmaynotbesoclearwhatthepatternisinthefollowingdata

#>0.0000.8410.9090.1410.7570.9590.2790.6570.9890.412

剩余469页未读，继续阅读

ALo54

粉丝: 5
资源: 111

从电子表格到R语言：初学者的R和RStudio指南

A Beginner'S Introduction To Typesetting With Latex

Pro Tableau: A Step-by-Step Guide

matlab经典资源合集

BiffException cannot be resolved to a type

describe a program

import pandas as pd

This function utilizes the following packages: GDATA

data wranglers

GIN virtual

aspose.cells sheet

python tablib

https://*****/open-apis/sheets/v2/spreadsheets/:spreadsheetToken/values_image解析mei w

xlwings chart_type

Caused by: java.lang.ClassNotFoundException: org.openxmlformats.schemas.spreadsheetml.x2006.main.CTWorkbook$Factory

Bill of materials

vb.net终止进程 wps spreadsheets

sheet.setColumnWidth

matlab dataorigin

python xlrd

unity epplus

最新资源