apne saa GUWARAT TECHNOLOGICAL UNIVERSITY Seen V(CE/ CS - Ope Eve PYTHON FOR DATA SCIENCE rah A OharePYTHON FOR DATA SCIENCE Sebi Coe 3150713 Senne Cnty Cogite Sae E) Op Be 1 Fo ti am 2080 Pin i _ PREFACE ‘De Jogoranc fPthon for Dat Sec wl oe es eee ‘lt Oneness omy book on ror NRCS ned ee Sea, he ow est soe ey ect as Po fr alt So as td angue exia nde o h ct The ‘ost roi pl mad of panig va compen op sept en the mgoran wpe Each capers eel spun ih recy Meo, praca examples and sre pols Ale hopes he Ya a range 19 proper soe ts pes each oc 1 Dal pon eur see A se Th ndersandng ibe tse concep ‘The bok ot ont cones he entre scope ote nj it mn the potophy oe atest T ake the uderandng she ic a cl ah aes ‘ee inreting. Te oak wl be rer tno cit these a ae he subject aces. The tens he fo Oi neting and pha acer Nothing seems my an ow ed mai oa Any gon fe the ampronement tthe bk wil Se acknowl! ee we Ate 2A iowSYLLABUS Python for Data Sclence - (8150713) ao |aie ‘Overview of Pion and Daia Swuctares (=D) to 1H) svodetan af ton cent 1 Feet Pyar regan 13 {22Asamaps Oeatoage of Pon 1a 22 DMA Tp Va ane ae 1.4 Pyton Dota str, en T sear a0 ta ‘Data Science and Pi @ 2 Dicovering the Math between ta Sener Yt nn 2 Gstinghe Ont Slane pine aad 23 Understanding Pro's ale inate cee. nnd 2A naring to Use PO Fk 2 2S Python's Cpa an Wenders: Why Python. 2 2s Wortna wi bon. eB 216.1 Worig tthe Command Une erin DE. a 2.7 nee thon orem for Daa Sees. sed $3 {nikon Soe oa se Soy pat 2.72 peering Funata SerComptng he HOY. 2a 1 Gog forbear wn an Tro, 2 282 Penge oa ine at, eu 2 ceneg Gt wh eva at 284 Pang HTM Doce ng bn, aus ‘Unit- Til Geiting Vour Hands Dirty with Data @-) fo @- 20) 214.Undersanding te Tol: Ung te pte Conse, a2 2 Ura eal ae LA nenareg2Onctone ve 2.15 Pring Meas rah npn. 26 £2 Wertng wit Res Ona: Ulsan streaming and Sep D8 on 3-7 33 contoning Your Onin between Muy nd Panda. 3 s.a2ceeg 0x ip an a Pn vas 3.33 Mansing 4 Cusing Cup arte, 22 336 hrdng Beans ae 2. concent on Tortora 2bs ping ma: Wor Wh HIM Page 3-19 {.sastmmingsedRenovine So Wert aa 2 eid epi xressins : a2 1.7 Ung te Ba of Words Mace anyon er) {73 Untangle Mt 3.36 Tait-1 Date Visualization epee “42g infermaton starting wth raph. a2 ‘..4 eign pence ering wine Sy as 12 Veg 088 aon 10 12 crs compron Br Chars on 42 Sowing Detbatns sng stapes. aon 15 oveng anced See as 47 Ping Soogapiel at ey {Vi 8 PN eee nn | Uae-V Data Wianaing SCS OB S52 Pn Wi SRM enn 8 5.22 bing eleanor Ota Scns ss 53 Peoming teasing Te a S5 53: Uren rnin Se SS Rung Paral on Mil Cte 5-8 6 plorng oat ara : S10 5. retng Aspe Vso fr ER S17 Sra mpacng colt sam 5.8 Understanding relation, 5-20 5. coniering the ChSqure Tet for Tables. : 5-2 ‘Solved Wedel Question Paper SSS TVW ED)— oa Overview of Python I and Data Structures te ing Sn rey Tap Se, Deana dees a Contents 1 teen Pe 13 xt ad FonePen See Bb rroducton of Python 1+ Dyin a ghee sping langage which can bed fr a wide vty of ‘eprom soem into inet eed wks «tytn isa tr jected langage, an eval on a wide vay of «+ Pyton ws deve in the sy 180i van Rss thn at Cn ‘sted and area CNR ign, yt 30 dn Yur 2008, 1 ytnontanents do ot ida ed wi spc chances 1 python tes on motes tis lcs peograms which dete a ait of esos a data pe [Amodule a le eta Python defines an sens, The fl ae is thermody appended Wana module te mode's name (a ting aaile athe va of he bal arlene «ame seated dit howenes hve gol alee + Notsies an saniin cle sateen ase fom deinitons Te re EEStor eh mle ae ncuron impor ene Seafthe eis ected aap Inogoet DevelopmestEviannt (DE) the asc interpreter and er ‘cern ar youn slang ws yt Tal bd an eo ‘Sr ceing ant moving programe tana necting Peas a porn debger A debugger pies ean aig contol athe eso ‘ra propam iin fining opr ers «yom mont comely tld by ae on epee: proves te wor eal ay wo exten interacve moe The window Dat proves tN + ythm sport wots: Nora ode detective mode {Normal me The normal mode th ede where th spied ae ished. py tesa inde yt nerpoer Tsmade slo ald a sit de + eucctve mde ia comand ine sl wc ied foie or ech otonent wie ueing ply od etme ve ems rte se 1.3 _ dain tp mba + Sar hen ciety gn ih pe oe + To aces the Python shel open th trial of our operating ese and then ‘ype yen. rent ctr ay an th Ft sell app $m ptt hr ee err am ‘+ The2 nda the Prthon sl iy exe and your omens ‘oho Python terete The el imma dyed on he yor ch sonst yan neers + Forexampe print thee "Hallo Wor we cay the allowing In scip ma, le mat be ct! sed oe eaing the code et ev In nerve made ret eae! nent ae psig the ste. Sema ane Widen fig oe Tt HED Fosters of Pytnon Programming 1. Python i highend, etree and abn siting Ianggs asl nd ey toe. is porie yon fe and open see rogram nguags Python an perform plea ig i fe Python can rn uly on die pains ach Windows Linus UND saMacrna ce provides a vast ange of rare forthe var ale ch a chine leaming web developer nd alo re ping (ERED svarnags &cieadvartages of Pyton ravage P08+ Minne the net etl and matan ode 1 Mulan bene 1 Langecarmunisy of es 1 pterge sanded an ser const Meary Dienrenage yon Pepe a roe slower an comp anges 1 enti th plas iowa typos 8 Variabio Avia way fering oa neni ope 22a le yale ame fr tsps ooton. Tis manor oso ‘Ari vlac Se mumber st ormore compli IPE Rowubetea name tt reo value The ol) oper we “yt yes sae Nbers Sigs its Dison Tape and is 1 Pht cn aldona commands dee «vase son ah alo ‘ig otal ecard 2 | Pv ect asthe 29 Rules RTT Spal tras aot allowed 1 Natblecan on consi lp tame haracr ad nes Svat ame sees ith ate ror wires «+ pylons uta ash value tome vablsina sige salen wich {ao tnown or mle sigrnens For exam ‘yxeyerest Binet 231 mre datatype les Fe Carpe Number nd Bolen. 1 hagurae ona repre wn rer vas Fost data pe wd 80 ree dal pt vas, Boca sed for coil ouput nce te Sepureboclen er ran fe a ET ea Corder eating ample 3 1 dat pes np seo ieear 1 daa ype changed to ost samy" and now til being Tyo sutnaly scare of py repent te eet a ‘ype ean eer als wile stared acento loon an lt nt Tiger vl cn be yen ach at ete 1,229,985 yon ‘sro resricion nthe rth ofan integer Ka ngs i. at: Pt suet ste Astin. pit umber Hie 32, 459,20 is scat pT dein points ‘Comple: A cmpex uber ois en dere pa, xy whe x ny ott el nd maga pats rsp. The comps mambo ihe 2.45 aaeade Python’ tse ramber tps sopon the noma maha! opertns. Foe natin the sp () pedo dion ae) ipo, anc wos (ree or expen ‘Theopertorauppta by phn at flows mae i onjecs ona Functions In Python, freon rut one ie yp of ec Some facto ae design toretuma al we ote reds or oe prj, fanning nt of cole cong ee of tees Indeed ender sname gion sng th “hyo Fantom fe allows cana | | 1intr Sse ni Pena See Tare wotypso ancien Pon. Pain Fae Thee ae the bin intone of Ft ih pe eine cna. 2. User elnl Patone: Those ar the fron dnd bythe ws Fer ‘te reglement at ire age How dow faction wok Python? cin re == | faction defn in Fthon srs Wh he eyo dif allowed by the cai Sct ete on mt te doating by fart rae — 6 + maple For ample tte nT nig te an Hato) + Oupa: ‘The doting oth function Hallo tHe Wert tes nooo vlen soared by omens Ug commas lets 4 tingle Tw daa vane RRS ae intro ce tran a Sens + ample tof a sum of wo mumbo ing atin in Python Gc MARTIN mu sstenrersear pares ees so OPS te HERI svinge + Sings ara econ of characte which ne ther ope ay sect ptm progam. + Stings ten ae a! ope unicde cae aes Python dw ot Iavea darned ye ssn cacti an onside a sing 1+ We dence or dcr the sting vals rie nl ts dole qu, To ‘wei evaterina sing we ae thingie and ure baat +The opraiar 6 krown ar 2 apt opr at te pein “yin *2 sms Python Ft * Stig eae writen by eros sequence of characte ng utes (thn) doole aos (Fhe rpg (9a Python) 1+ Fora ings const of ambi character uta amar. Char it tn wich want changed en he fat ng the et tie shup et nee hf ing ey toe up 4 ample: “My names Somat Rakes) + Teresitottisietesnings "My names Rabating Operators 1. coneatetion Te aiton ptr) heen he olf entation when atncngs The ret fang te gs bree eng ch consis = Eomele + Nospaceereter tracers btn cnet hg. Pym lao concaoate stng cosas that ae sept by epee ve 2 pape nti) when ed benign a ge rete 2 ning nih he id ng oped he valet hen 1+ nape flowing sae bt pit ve hares Pylon sport indexing end sig operations Tost asingl characte fo ing ft the sing with tend of the deed racer surrounded by ‘ume bct (emer th he et racer oa tings index 20 EB Arey | "+ Python das at haven spp fr Ay Pyne ed sere he 1+ Asotin at ite aaa ant isin he ats ype of as oe in 2st Ory the seme typect dtc be sre a 4+ Tore sayin Python we ed mp the aay mae stays, An campo thon low + ample aus + Line provi main for sing allen of cts inden by uteri yl ir ds ve, meaning tat mes ae ew ing, ‘+ ach lane be aces by a nce Nt that Python ins tt wth 0 sts ae a mute clacton of ten. We ue aq bracts fo und te temeth Themen ofthe stare vay, ay an be mes sting tn ae | eine ej or even ta tym ciples data ace ey singleAa fit omen noan ante Stas 1 seit of mumbo in brated a faows “+ Indentation vry imprint in yt Jt he tee cero > nae an Inrative pomp nto the he St ae Pythorsprmp or mail| ee Aer sing st in Pye tie oslo seach, dad emove es oo) thet + ample: emi io 1 tists arate ety an Bees 1. Asie ee ae yg otn ofthe daa eit 2 Use teil ano ang or append te ats pty List it that oie no ements ed empty st an be etd swihenpy bre + ample: Aging Lis Vue Vare: + oust: Thee are no ition on th lame of sts It can comin eto sotee os) sista thas aaeh yl HEB wie + Atop a egaens fv The ans an be any type and ae indeed by fncges let Tp oe inte ‘Topic oa contin data tat are ae trots tose pe ikea Pytan Tat pe appre random cleo fej Tp 2 ed length be, “Tople suport reing fim tac atin cont an is. A he apes relma you can huge th ae faa wot making 2 Ran A ot "Cal> Run Al Abe toner xc ate Ei Restoring checkpoint + ney Neck ocspas pd sv yur hci you cose the ab. = + Tones cept doe the ety fund on the Fle Revert Cine ‘+ Tismen mas ppear that yo in have moran ne heck fle ba he re ever ethan one yn 1 Tig. 314 shows mena orton checkpoint.| Ss [ibd retorming Mutmeia and Graphic Integration + tytonNolcbook upp otha ing plitfrn anda rotation pom. etna it nts ncn you ra Seapine ny los yOu cet appear spat of Se omc se anata ili = iy ws he agen ot erative vores wel + Fabvng ig 314 shows example nbeing pa F314 amt nt ne TPE NSteaDol «ie dy int seed pom os em et Pox Notebook resents yeni awiiwenetiaae et 3 Sintra nto mete mane te meses - 1b) These eppresentations are stared in the Notebook document. ©) IngunerakeNokibok wile et orale reeset, 1 Towa nig the aged 1) tea tyme pe ee Sig mmttchoneer Bl Workng wits Res Data: Uploading, Streaming, and Samping Data 1s lord in al campeon ace te a ble a. Nomally use does not ints with sage aon. Bt data lad ‘memory fom sage aon then proces sare 1+ Scie ary sts ty alin date cn conc as ars an ‘os aces Echo rps cation oars a can are 1. Upoating smal aout fast ema 1 Conveiet meth wrk with dt when tay lo it ln memoy Comiderting e + hen you ln 2 lhe mie dat aalal al nes Sy ang eth carte cd pe star th pen) eammand om file act. Te open anon ace the ‘eps alan sas oe It prom he action aor fo the es ode ter wie 1+ Tce dation om he ayn fe mmo. 1. Seaming urge aout fatto maar 1 Lange dt scan nt belo to menay bs ey si Remote dla sets las slowy bese of eteckconmon, Sing some ath eee ‘by mkingi peo wrk hth dat ait at tine REO ie| 00 so a0 tn vr 2 te + sean dit sng Pytionaflow thon een 8 Te open ile lot ite iter the pen le. [BER] Accessing Dotan Structured Fate Ferm 1 es ata es hat coin rth. atractared atlonshp ewes ‘tered Aft pee net il of eo werk Wh Arcane a linet faving SY, SV formato nay le format. 1 Con Seperated Vue (SY) le whi carta dt vals hat re pratt {Por ctample 2 le sve with are “Dats “CSV fora wil ape at uu By ong ae een we an ely leita ts "CSV" an dt stad aba format The CV le sd fran capes ute sine 2) Nbr Snes acho tees 1) el spate ty commas ©) Record pate bynes 1) Sng nsdn dbl quotes 1) angers eal mbes app tout double gute: Alcon uch Enea inate at SV 50 ht they Bee “an Bet ec completo parte dt ils ano prvi weih soma abot nh i El aul ose ner a edt Ris an XM bse fe oat ca By Msn Exel The XLS om 8 Ine with Mec Ot 207 ext dita oxpibad under he an atm in sh. ah XX le ry containon reste So wohbok cn cos mile sees “Lert hed om XLS le and inet het ame Fo ing the ata oven se the Fada rary eho STi eta porn tintttya EBB eating roma Text Fe + then provides the icons torn ta fom tet fe: 2) rad Th fron et ys et et ae he me ‘necator eff no mabe pected I eat ech net eters hes iene oe det tre, rine Ti ncn lows yo td i Et ©) relins) :This ncn reds hcp infos nthe ee ‘dou her bat dling he and Phan nel Me flowing rt typo aC le 1) Destino ean ned ete ©) A cage wd 0 septate tal ames (ome, ab eo) 4) dtr HTM Spal Chance within eB wid okson hearts [EBB senaing Det in Unstructured Fe Form "+ strc dat fk ost of er fs. The le done spate the is frome other ay way “+ ‘Shama provide enya uncon lying sng ages [Al of the popular image formas, such os EMP, PNG, JPEG, and TIFF are sipped. 1+ Cmsteratloving wooden ie RRR RR+ Avinal epresenationof how hogy ord a ony any as flows: Te fp caiman [EB tanagng Data rom Relations Databases "+ Relaonl dab sssinplsh bh the manipuion an daa rir tices SQL perm lst of management ks lato dats eal abs cose of neo ne bl of orate os the fac are lled ord andthe in he ae areca Bel ates. A {ees hat coin roe led aber cea elitona dobbs en woking ons data ce Fo, 0 may wan fo amet Python ps ‘wi dats A ry Snow 0 SLALSenybige the p bete SOL ar Fron, «+ Tei hing 0d sto cet sn eg wich an je at eel mage ‘ensontaa tel een emete ‘Seguro ng ronment G6) + Thm rel fomat nt an gee ‘Go cet pore pens at Tesh ary proves spot er SQL stabs ke SOU, MySOL, tyes an SOL Sener nero wit Dae om Ne SOLD Te Not ny SOL (NSDL) tabs ae sen neta ogee in ‘hich elie lan ec vei compe + NoSQL dtc prove fares of rte an og of ato much inn wy fan nr rato database couse. Tet gta we to i cre sah a oman pr the Nong se. Well se the fo cine wih te runing deans ‘tne Ue te loving de odo 0 Siormieecd El condoning Your Dat: Juglng between NumPy ad Pandas sy Nay the core Hy fo set eamputing in Pt provides sgh pefrmaelisinersnl aay et a toler working Wit the ays «NP he fulamental oc eed sete compting Wh Pen 1) Apert eno ary oft Basics algsta none Bsc Fore tasarms 1) Sophie aden amber cpbicn 2) Too forint etn ode 1 Tock x itpating CH ce “+ Nady ean ener package to Pon fr aray progenies "doe thc are epithe non has Cpe reas a highseve t maiplaton to devlpel by Wes Mee i ‘atten the Numy pcage andi ey dats strc cel he Dita + alarames oe oto tran mapa abs ta ino eras vss ort. ada i bt on op of he Nu pacage ening of he cae of Naty esd o pin! in Pandas, Dat im pda sen ud Sot STR a SR RR Rastial ana in Sy, poking nce en Mao, and maine ering gets Sit. anda th ary Sr dat pln al api Unseen Pin fry dt eo ta allows ou ar dat mo maple rc Poet ising dll your data reap erg and jan fh da seach daa, rope ce FER Ferng out wnarsin YourData + Dupes the pclen or data cece projec ane ge then Priming dope dataeanewaage ie + Fin pete ipo ce wl se tin sp ala el howto sy ad eel you cn remove the dupe data ang op. Sept) fantom pds + Coe Dame ith Dube aa n= ft nee rine til. Sper eee ae Ice iaa 23 ‘Taso 284 7) at pau Su cimee = hate nn.) ikea, Seton + Drop dupa i We fst rane cake, bu ke the ast beatin Int aphened t on RTT maa EEL certing «DataMap and Date Plan "+ Overs of atte en by da map, Daa map wal fr ding poten poten dat ich onda vraen pombe err ising a en Try ting Fon sc tat convene Fon stony se a Pans i = (m3 MH EE ‘Siren ind ua a sr, [BEE waripuaing & creating Categorical Variables + Cotgol vrei oe at a peeve Hed ection of + Caples in oly ae on Het and aly Bud, uber of rable valee For camp date aot fermion elated stn exw eel Setar he our gee age opt Attire f the data you ae worn with aed prot Ju wl Sea he rotact ype maracas li 9 + Method for rating categorical varie an han ig cack wher ome ‘en = paca ae penne) i pint cna a Have cyclo a aor vale, contains the vas Bla, Ra and ET ieEEE Renaming Lovee & Combining Levels ats fae vce rames ae yc wd ay times when ranging dat (Cooder the vrs rahe Hee 0 woe and rnd rag ogans + Compara ata has cpr and» ade poet, whch i ti posible als and wher he nei ters oo + Renaming capris i done by airing ne vl oth Svea teas propery oy wg the Categorical seme core) mat Sen sw ng Ya a (EE Dosing wih Dates and Ties Vato Date ae en provide in ern oma nl i covet i ge forma Datei jets fre aa, + Pyton provides wo nthe mating dt an ne 1. tr) i tun tine vale in sing whut any oat. 2 sind faction it define ho wer wat the tine vale ape ae |g panda. atte wth te 2.Uana andi. detain with dat ad ine mai rpms NS ec cont Sing t dato ng spe) facon Ths frcton alae nue and tne male 9 pare 2 ing dain ante ec mpl. + Python spin) i cats at dts de aya‘+ Time Zones Wit atti tine zoe erp by subse of, ‘Se tin i a abarc ec yo te o tine 4s ed provide ppp planter afew mata ak aa ET Finsing the Missing Data + at in have ing vale fora ramber of rs ich a sbervatonstt ve ead an dam corpo. Handing sing at patna ‘Rony machine ening lithe do ct spot swith sng atex + You an od he date a Pandas Dla an pit ay icon sch attate Semis faye opel Pandas Num pd Se Leary, weak mining aha [NaN Yas witha Nal ae ged fom operations ke ma ont ee "+ ethene tec he ing wales Panda Date prove 3 funcional) rete @ ne dre of sme wa ling trae contr ny Tro Fas ely With Trt he lace NaN i oil date toa acter paws ncoana mesnaene Teflon sao NAAN vl ig speci! en + Symi ‘we 1. vie: ava ht uo el aes 2s tke ring ale orm eye np IT ls vas an ey plc. itis an eager ae ht mir he mh abe of ss forwarded NaN vale Se 6 doncaster hat donne ih Foti [Bi sticing ana Dicing ;Ftering and Selecting Data + Selecting ig ves om 2 at ut ne ptf the ry. also ose to sl your lis wh meas ting mail dement ro your it. ‘Thee fant ers se jot hat an we sd ols i ope ‘6 Tesco df ie a given soe (tig, es pe, soe ang) otany bj Whi sopra Prete When seeing date wit yon fs pert remeber salons stat wih 0 ar sccone of mie cles ae spate wh ca (easy. beer execp the ey Lt mabe a [03] wads ht sed id om at ot he th, 1 ing The at a ng dataset nan por sh als ising suc at yen ups dt edge + ow 0 dete single or mulipe rows om « Darame objet: Date prvi enter nn Sop) Tt col ingle o it of bel names a date the creapong 0 columns As delat alu or as fo dropping rw ena ot oe 1+ ko, by dla op doe ned he iting Data, ts ew duane fe want ape teeing Date im plc ha we ed ‘epost ate HEEL concatenating snd Tantorming tn at sconce projet dita is colt fm varius etn with mule lands All ltrs have hr foat Ie iff perm dts seals uch parte our th ay 04 ag nace nd eitos "+ Dat cj provides acs th ce dat an aie inion contin fn dat ad allows you real am th dat, Ad te ce ey Sing omy ad new ails and ml proper fot abl‘Append Method: The append method in pon as sige tem othe xing Tet dort oa a et of en but wl ye gi ety dng Aer aesting he ret pend on hl hit of ine by ne ‘Te append met bea gle amin prmeter eda ht ote uf the Int To ame nde 4 Ut can be mums ings ances et, + Tapper math ny modes the ginal it ect rum ay ae a ua a lst moa rent i + Soak append ented ton re fr sing ania, ang ew lamer eit Bat ey ve ier EEE sorting an shuting he shu) mao taka agen Ot ing tp) and emis the + Wits Data you an sry ines ihr i: ‘sbtrtane= Deanne emma her adh = ‘a eae ate 1+ Tos te dats you ue esr en) meta tie wh ceo se fecinoung proses You ca a determin wht inex aang econdngentr Pyro Sie ao nvr ae 0 oe 1+ The random) mato wl sls 4 tin Python, Bs ea ‘molt ore evry tm n= yan The dul cn nol wth ak optional eo pramete + randoms prot rand rary, which plein pean fumber gente in Python. Ths tans dat befre we can te the ‘anon method, enero pert he anon Ua ro ou coe We fandosoby aiding the loving ar ade por random 1+ Theat othe andor) metodo flows + Trane at) hi a in parte: 1s. pame elit eps rotng whose ode you ato she ast!) 2 ener thename cfs faction ht ret am at betwen an 1. ante nt with at ans opon EB stoping Data: Working with HTML Popes ‘PEM pci the cotta nyt of wb pages, The cnt cot {ih om ings tik, oat fo sach engine ee. The yeti the {oem of ext oat ground a me HTML a wed pea ink nd 4+ Hyper Text Makup Langage (IML intend as 2 omen mau fot png eter rat er wl dfiet ours 1 HTML documents are the Sundand Conon! Markup Langage SCM) documents th eerie ht apopte repeeng a trom wide get ppt ‘EI documents ae i plain Wot forma hat contin embed HIME gs Document an Be cated any tet et There te ao any ote ol ining eto, Seige! spect 6 ait eating HTML damon To PET dese the erucatl ements 2 document sch a bade, and ‘dese, ly iron and Se ilies ote withthe ‘hit prove hyper tot inka Webpages were tenn HME el om aOR emaPenton seoce 8st tt a, (Bi Prsng xa. ana WT. +0 wands for eXtensible Makap Language. XML 2 mata langage fe Cent conning scien XML ia set of aks or Sg ‘og ants erm, + Deine bie with sel te HOA for ube, soe ino Bone terse etna pans dt ane of consents nnd [HERI Using xP ath or Data Extraction ‘oth pees path eprint males XML at by gating down he te, XPath a ae eb quar language bth XQber Dan ST + XPath prods cmon aan and seman for tray sed beeen oUT ol XP Ph ed in Stadion tof inormation nan rene an ty nth ts 2 cme dn ee dn XO + Pass artes by paar oct Tha i separ fr pring bth XML see hoes) IML BEB] Deatng wit Unicode Unio is ot ance codes mare ke a i dain of hrc an Propet andes ano some anys ntng tut how ose ee charters equ orc by cmp, 4+ rico dost work em af hrc wen ef hing ele ade reins. : + Tyan 2 comes ith no rend fot at cm be wed pee stings san ote + Since Python 30 tinge ate ord at Uni nan nance of he pe Enc stig o he ter ar opi ary atin th fom stances oft ye pe EEE] stemming and Removing Stop Words Semin isthe ces of ecg wordt rt wn + Tokeizatian The poe of enti ein wed aes rents. 4 Seming: Racing td won oa cement, 4+ Remora of top wade: Hamoval of cominly ed ds sey be wef fereunig. 1+ Theat of mening nd evn tp mde spite et an cs he namber ff element hl lth nal ners ema Stemming ia flow Poiana bt tsDeena os sae a eRe ee ng TE le HE tei Sarees Ea em ‘nay hese) tg en eg ted an a 3 a mt er i tt + Removing op words sop word’ ae cnony ued wor at re way have any ta sal guage rung Thseincdes words ich" ‘tncom seen Tere ee niente ng =a apne Ei nroducing Repu Expressions + Regular epson pce tinge ned tes for mating {eof ings The Fionn provides eur exes ppt. + tntyitona pula agen serch yay writen. saeavnneaoees tet oe pa cH Dy a, + Te rear method tls rear expen pte string anders {oc a pte within the sing he seach ern ech er ach cictrNene ti non” arh a tg vr hn a [ilusig th Bap ot ors Mode ne Beyors + ago wording fea mate whch term ae amen ad he hums ae weds Te ae mnt docume with wed in he ma toil be cunt a word ccc thn th dare of ih + Cir ond nh ag of words and pc df lest ed eek words downing “Tao ow document te pete fn NLP as a tapos This ne tht choca snd a2 aeengh ver ith gh ega tothe voir on 4 och dines of his et compen othe out or csrent of ord in Atcumen Bega oes vate documents o Seth vectors ‘ke ho ore mem se with ng asi oF Machine esr OL) ode and a +e 37 dws emigre of ed pretation a Tigo nt i apo wed ern Bi Working win mea 1+ The serial cng test mnng ie rani wich ea tf coset oF nts sequen ens ram wens of ag esto sees, Te em erect be wed etry an sie. +r issn ea uri ae he ge words pret in he eet Siena the eumbiatn o 2 wor Tgea(ra) 3 wots an + gram ie earaos sence of ten he et ou want oasis. The ems are phone yb tes wads fb is The rerum ees RR + Treen generate anion EEE. rptementing TFADF Tranatomations “+ TRIDE sans fr “Term Fequnn vee Data Freaeny” 1+ Tem Feeney per the ng of he word ineach document vee ete Fegesey i) ued to eat the weight of rae wore cea rm ira rr en gC cine ‘foment eeu ipa arr oor mgr toate + Bample ‘nt arctan npr rae SS lhem minster eet mp neat > fen ores pr Pe Sos ett roo, 2 Ra aterm cenntiietes, maanssa oer reerpetro sme sm tg Yi 9 oe Smee BB erste te cay ae + escent ee to ‘heen ten as pn rescemec= er “Testi nlp erp tment eesti) + Seater teaey non tath snl nr al ape Gromer cml nce + Sayan ad pe "et dane ct ren! nl diac eo Lt ie le ‘ined *" begat gc pon nol ays Nee pe may ger ‘an nts ad ene apo ae eee fomesstens Spenser nant a Hak a RAL a4. en Tne DA lea ov ream IV Data Visualization ‘Syllabus ‘tng are Sepia Cap Bn ep Dea ate en So mt a “Sig aise Se en a cata Be ‘apricot sng am ts, ‘jn irene Sp at hone ph Showin f ne i pi chars. raison wth ‘roe Sime smo iy hago pg age Sy ‘re Fe Si Ep ek Pat od 2, P| Spe eat Mat Gt Rey oD ok ‘Ser ire ane in cr prope deny Sm Dat war Saga ek Contents 41 atang mato Sag mth Goh 42 Vtg Oe 43 shor Dut ng open 4 Dott Geir vag Bors 45 Sagas Pat cr Sette 45 psy Tron set 47 Peg Gary oa 42 intang Goes[il Vievaicing inermation: Starting wih a Graph + Data iain ete pian of qn inoraton in api form nother word data sultans tun age red eal ato vals fata cane rte an ban neta es. + Good data vito are cot hen cnimuieon, at sens, and dig elie. Data vss dove night fer Hey nhs ns cmp cts in wy tt sce niga an ie. ‘Araph ie eply 2 rina representation of mei dt. Mato. sports gener of papa hat ype + apo spol Fon pocge wd ob lt Mati can al be sod make 3 pls and ination, + Line psn tn Fyn wit Ma's ppo irr. Toba ne ‘ou fit import Map es vst cornet 1 pet Mao yor ap To deine pit youd sn any he mah ppl dl aan hat ou want aay. putamen re ow "The pp wil “an lt inthe aground ring #0 he sce when me ren. aerating everyting we nent. + ts With hath graph sald pop poo sreties tea pop ude, {youay hae ten an ere Your rep shook ook he 1 Te window ia mapa window, which allows uss our graph awl ae seat th nd aga (EBD orang Msp Lines and Pts Mar hn nee can ben th pt. To dd anther line, ut lhe lati) fanaon pain Inte nape wehae te ern ves ry?) at re plted one ectar ote ey ‘+ Oupttthe shoe code wok ihe tis ERE serving your Work To Disk + apo pltcan hese a imag Se ig te psec, 1+ Te saa mato rage a fname be spec! a the Bt argue Ts ‘enamel en ln re parts exes i esi no etenon provided, te ogra ae of ve frat sed itn, 1+ Te sot aio sumer of wef pial rset: 1. dplinbe wed ost the reson oft a amar ae2. wanparent can be Tr, wh cas the aground of the cat be ‘epee, 5. bowie cane taker he sinh ang ox tsp) ran fe ouput image, In mot ces 90 Beucing box Is ested. ting tons is 4 rttocince tog tthe pl Jas option pce he mount of adding und teimage ERE seting eA, Tks, ris 1 The ates define the xray lao th pape. The ans ans rz, and 1 Anand to lt yer Anica be tought of a fx anya hat Src al bts a dana An A conan dager aus as bes Sun auetsinoe « Retmenyase wre eee pot sina eee ee oy a, SH + A.gid an beaded Mapa plot using he pg) command By dean, te gid tamed of Teramoto ius 1+ The ony alld pions ae leg) and tae) Noe Bt Tre and Tale ae cpt and arnetencned in gute {ERIE betsng tne Line Appearance & Working with Line Syl + Ue spe lp diferente graphs by drawing he nes in vrous ways Following in seis ety Matin Peis 1+ Maps addon paramo contra i color an yf he pat saan 1+ Tas wll mae the ne gre, Yow an we ay clu fr gr, He agra, eliotn it char fhe lor name ‘tower ne for bik ean), 1+ Youn iio where nee forxanple te dais ~ make aed re Te cane ed ade othe core ee 1+ Youn seers nth dl fo dato i 0 for a dot ln He tan nape1 Having int shped mash ia et yt deg tem lent 0p of ats pins If your core gop i al dee and your eprint ‘oup all Xsthe tere pops ot ve coon viewers. ‘ew 9 mater) Ere ids ey + Tensno way to pc mull marr pena gle Kt cl Bu we can septa car dats Ot nto pop af pe cach mar se spe Hee Chopped cudnt pins tree eq oUF (ER ving Late, Amotations and Legends : + To ally docuner your graph, You wuly hae lo est abel antatatons Thlogend ch les matt aes lw ERE Asting watars 1. Libel: Me ey Kt veer tow fe sae odo data Mars pc yb cach spon ie ph Une ese tent £ we'Sie maton end been pl ay and 2 Amon: Hap eed he ewer one fe di bern cm Stop wey 5 ney, mat et have fr he me Wok ae 4 agen: rect mae con fie da groper otemuter + Thefallvrg eagle hore how ait ee ah TL ragotapt v<2* weneniusanaasinen Bee 21238 Saas cadet yb, pH eee RA REE cma Sa RROD ine neonAo anet8\ an leer shows pit values wha he mouse over he raph Dt | nao were + ling ple shows ond action oa rok ptt ypc asp peo an paow x00 ‘nb lh oworpenrn roots) paste Prawotone) A hoa RRMRNETENTE TN omen Tho rst pin stale fr cnn th appre altar of tpt lped.Byefatthelpe vay apes when ther te miple eed nly pee eer Wie hw ingle ote. det the 1 Alepnd dnmens he india lame of lt ich ok pana a2 tbl hat coins orisha poole nf tween exch ne. teme= ted ‘eves Pt > i the sp ato we line two fn: quis a ce ang yan vehble Net we ft plot he nue finn an fore abl pant Pas the valeSqsr Funes, ‘+ sb th vale dle nthe Ml fo ga fncton. Not pt he ‘be fanctn and pas Cb Furth el paneer + Teopuioks ie teWE Viuatcng ne Daa "A gp of rp a which 3c ie int srs tht ech preen & roorton le Each torsos th ae ach ae, hp chart day dt inosaton nd atic in an easy rea pel eet wh vying ss teling how mah ope dt element is 4 Fig 2. shows poche “+ Pi har ale kno a ce graph The gr test sore otha ea data wo ter. The ne pl cats show comprar + Person te dean ang he acon pe nthe ppl mode, he aw tion code camp rae a car sgl fai, + By lat tp) ton of pyle amanges the ples or wedges in ape cha count cts ction bamete scone tenes ‘nga cin ap Ms ‘Seouen mches an) rnp ene, Ute pt M1 aH) ‘tate -npn mance apa ee reroerrry ‘+ Te eset pt of echt ie he vale. You cul et ap chat sig juste alas asin (HERI cxeating Compariaons with Bar Charts 1 Abur chris aay ofsummasenga tf tego dt, Thebes ha sing» number fbr each repent parc ego The Hoh of ‘ch barlproporonal tsps ean +The cegoce old besametng ie an ag group ors eograpial aon Ite leo posible cro opt eat a int another eee cok ne dt, hich rable you to eth cntsbuton am eet aero cach Bro [poup bari bard + Bar chansaecompaing vase b=Bilstoning Distributions using Histograms + thahisopan he dts are grouped oan (0,138,229 and then pte (homme tors. ach Sr reper ange oft The ih of oh a ‘pura fo the with of ech egy, and he eight proportional tthe feoeny or prtge tat eg. + lprovea vl nterpetasn of ame data towing te mumbo dats nt tal within speed ge of sad "be 4 Fig shows htop, 42g 4 Hlogrms can dpa 2 age arcane of dats athe quency of he data flue, Te men etn of te dtc determined y toga, nado ten show seater orgs in the dat + Mapolh provides ¢didatd incon 1 amputee ipay stg pening + Cae fr cestinghtgram with randomize “tem neo ne et ‘a em Wi bepicting Groups using Sox Plots + Box plot us Boes and ies to dept the abuts of or oe group of ‘med, Hox plots ae ued fo show dhrbatone of numer dat rl ‘specially when yu watt compa ha even mile groupe ‘+ box plot my aio ive ines ed whiskey, ining dats cite upper nde gurls, ‘+ Ines cst posse toa many Fn cts bt pandas ae Pendle tect the sect names wl be tan! fo ena the ae ‘Aaa, you ean wse Categorie tp fr he oping ve fe corre theorier of petlaret + Tefatoweg eagle shows hw trate box pt wth dazed dt ‘oct mepran pi poreeseraenir«Th ll bongo ny dia np Alot armies have eal stings [Biseing Dota Patterns using Seatrplots + play lle fal be ot rhe st of ata nied ey ot ae bole ter lot Yeah +e weeing wth siti dt ti fn ber! at hare ze ooecons ‘eesn nt of dat or example te mss nd ght of pears rei the ‘Sorte pen the geri mss «Tot ut wher ooo toe fa ae cnc ter gras in be [sar diagram is oa for aang ron between to varies. One ‘eles plot on hos ard the crept on te verte «Te pte of hrnersting poe an opal stow ear pate CGomony sate dam i ued ts prove er dapove comand tert ‘ester gram shows wong, dor na by tal prove at ere Novi cer oer. In addin to atowing penile case and efit htop ect harm can sow tal evil te om = ono ‘et inunlowncrtttne varie an be sed sa stops forthe eh + Tefoowing sample shows ow ots ate sepasapt Pirate toe eNom Sire renee Peering Saint no) Sires ioc + Teper ative x tne pl ons ay, iho ku deen be col nan ype ater wes. ‘+The pinay eee of paar fom pt tht can be td et sere po where the properties ofeach nc! pit as, a elon, ee ‘ole cambeinvidany cored or mapped od. (BERT creating Advance Seatarpiots ‘+ Seer ae ep inert fo dats acence beste yc howe data ters tht ees her viewed naira epecmcunomitue weet. 34562.6930) ‘eto ss nt aero 9.4567.89.39 Poatomaee meas, tein on Oe) ‘seer os at ala) ‘arty. seroma soho) RO ROT a arrearsocr oSmse ewe ot tnt, ‘+ The cat pay to datasets We sinus betwen tam byte our the tier [repressing Time on Aces ist of le lt itr pt of its nd aes a Matt sing ipo. dc Weil ng Pyrat sind cad atte tet tiomdca) fr paning the der Sp, it ur crate 2 python Ae ale "Potties py and mabe newer impr 1» pen eco ‘Soomamnininrms useutecce? ime.) setemcecn 3 Seema 38, ‘Seeeaot 1 wemuaicaa sin a ” sete eal (Eid Pting Geographics! Data ‘+ The Basmap module nds ata or dawg costs sd county bounds + Tosetth ee iby defn the map bung boxing conta: + Tbe: Sting he bing oeTe 472 Set he bounding box gig th mer pot in geogap a mall Frei wes arin e-toc prea oreo ee: mane) Bartana Lae oan peters mamnin 8 ithe cues ne Eid visuatzing Grape "+ Te vnuslzaton of graph ha ron ob ey flo eploing sitesi ieee spplction dooains A graph Isa depiction of data shoving the conn een Sait ges 1 Grp fet fects we ae comma In ne way. The oe a cad verses or nodes. Pcoily, we small eth versa cena aan teen wo ver fey ae conte or rte Tes ie re ‘less. 1 Gap ntl ren many way a hey ave samen way to ie vaso tino or conples symm Computer Atenas la petra form «graph So do the Innes telephone able tens and nee ing spies. 4+ The Ward Wide Web: One an om 2 raph ofall web pages and make a ge ‘com Page At Page Bi te pti rom Pape Ato Page Bin ice one sh ose dicted eg ening ech eas a ein Wc is Perle withan tow 1+ As used in graph hear, he term rp doer ot fern dat hrs che tgs oe ber rapt nnd teers et eres (at pois des) rao ges olin) hate the vers. Sea ameget in beter eo ound ‘ere int we to or mere nes ast Tig 481 a venexand edging it vane tsa gh “wo vers ht ae cro ina eg 0 gs at share 2 sete ajc an eg and recone in date ancient Muiphs Mlle edge araowe eee verte Simple gph: A gph witht oprand with at atone edge bate an eadopaph A geph lu con mulls and gaphos ‘Avundiced paph scout erry two nadine etc re pect ty sme ph etre Components of raph recta th diet sina ome sa gps ‘dnd rps coment ung indicted graph connect is ‘ong conmeed exch cn ach ey er ede bya eel path ‘apc ala be a or eevad A wba t ambes tached ‘he net inate he argh oy orm or any ee ‘eter nodes Foresampl Fig 482 mit aor th ann rade nen doar betwen some ois +t Bre comes eno pointy hy ae aid fo be “eo” The two pine comeced ty sine ae lla ends An egy tiie or tint ‘Gen pot since” upon at pn Two pe tht har a ta kd ‘bein, 1+ Aabgzaphofs raph ia mel ot points og with al thE cing embers ofthese. The arp of F482 at des the UK, Cota an ‘Algraastvo ines (UK, Age) an (Alger, Coada}. 1+ Te dee of pit ie dete athe nar fie int wen th ae. In Fig. 482 the degre of USA 3 bce a3 I th ep i cl an lite ithe eget pdt. Ina ice ph ths hint na The ge [Anodes ecu om anther nei te xt path yng oe ‘+ A conned compo «main apap in wi al node re ecb ‘rom every oer Maina mean at ret pone bgp you coal tld str ade anywhere nthe graph th a ca be eo th Siegen al he nao a he graph oul le conc 1+ Ter dima gaps thr tong compos an weak components A sang ‘emponer i mana map in whch thre es path ran ry pi ‘ry pot flowing all the tein the dion thy ee potting. A weak Component «usin graph wich wold Be cnn re gored the deci of re + Te folowing ie Pytan pts NeaX tea ane Ph Srmamae Fetch inte aaaUnit V Data Wrangling ‘Shilabus rar RARER inaneBid reduction to Oat Weanling "+ Data Wengling he proc of rasforning dts rm gal“ form in «more dle format and emai mts fm vous sours 2 tingle coherent wef uth posing, + a waning a cle dt rng 1 The pinay pup of data wrangling cane decd a ote tin cent shape fn ther words making rode abe proves satis fo fare rowednge 1 aa wring coversthe flonng proce 1. Gating re rations suri one pce 2 cng th dat together acting the deri ting 2. leaning he ita ne eo enoeoig ing amet + aia wrangling the rose clair and eihing a data et lied frat rete del ang ine ine Thee aryl tere steps hata pt dat wring procs 1. ive Before oan ve apy, you mat eter arden wha ‘nyo dt wc wl fom how you warn aly How you Wage ‘tans dts forename my infrme by hee hey oe td ht {hay ugh ew prometenstheyreed 2 Srasring: The mere gas he dats whch ees Bc 8 la comes nny feet apes nr A single cl ny amo Sever ow freer nls, Oe sae ay bce. Movement of dhs made forest ompotton an nls 8 Ceaning: Wat happens when eronand ote saw yout da? You dean the dala What happen when ste dts entered as AP oe Ava Pash ot ‘rch des? Yu cen he dt Nl aes ce chang an anand ‘eematingipleeni tatiana ay | Enc: Hee you take stock i our dat and sate abot haw the ‘nos ta mig augment Quon aha dri data wring ‘hop ight ew ne pe of ata dee os wht any are ‘eh ter informatio aro my decom aking sot 15 Valin Vttion re ae epee programing ounces at wey itn coming, aly, and acu Examples of maton inte suring unarmed sd ste omy (Ge tt date or config aun fle gh ck ao dt (6 Pttihing: Anis prop the wrangled dt for me own, wether bya portale ero fare a dc any parila speak lage wed to wrangle ald it Data wraghng irs dentin that limpet of igh is pon te eas with whch can eae seca yor (Eid Poying wit See. Lear Sciam fe machine laming rary for Ft. I fates vant sips he spp yecr machine, ano stan eight, a i ‘homppors thon ner an cei ae ke Num and Say. ean ue perme inpliled BSD mee aes tbuted ndee many Lins darts eneurgng axenic and conser | Schaum sa tray 2 elon af le nd fc tht er ip Ino thon progam Using soktsam thre mgues hac Python Prong lage ‘+ Tota i ba apn he Say (Soe Python) ht ms be ill ere youcan we shee Te tata 1. Nunfy ue mines ary pkgs 2a: Fundannt ay sci computing 3 Matploib Competent 200 ting 4 yen ntan rcv nto 5. Sympy Syma athens *+ Schitern the pakage for michn aming a dita wn experimentation favored by most dataset I conte wis range of wld arng ages err arora tein roe (EBL vocerstanaing Casson Sika 1+ Scien ae 2 Ny octane approach o mache larg mode very a Siem cant nl ane Basler 1+ Setar fotze sete dates whch al hag ae il Apt tom esti the le rom whl otras nt, threo se ‘ype coving l he ac mace ering nce RARE a RRR wna aetre se es oa fe es Pa as Leming [Ezz tng Aptcaton for Bt Scleoe rapne Das ce pple eco dann end een pb, Georgy ates dese penis wed fgg ee retro soa ‘redoing tn ‘pop en penny teen cbr eprom pon + Sinn ks «hi bert appro to mucin ling des very Ser cr ht fn ln be Ban stinata. + Alec wit schioim shires form oman tae APLensiting of se pleases onetime or bing a ig als 2 dior ice for makrg prions ar anslomer interac for covering da. ‘ttre The eine leat is ot te ceo the Hy die ition inns of jc an ess it tod for ating model ining 4 AD apes and enero Jeming alge (¢g for dain tegen or hinge ofr obj ipl Ne etree tine ming tse ee eran etre acto imeriraty ect sal provide esau. 1s pier ince ete he ton ofa etna by ang a pei ‘nc ht se amar X et ad produces pects o Xttsed th Ind parameter lth otto ‘Inthe cn of espera! ring ein, he mati pay rts the die abel erates coped bythe nod 1 Srvii commono most er data efor feng 2 eng agin, sae evo nthe ay plement a Manor eae wc eis + ate aninpt sme re Xtand elas outpa nsomed eso of 1+ Tht Housing Dato const of i of ea vaso plas in Si ly Alone with ris, he tn su pve nation sacha Cen (CRM, ‘ses oneal sis in he ow (INDUS) the ag of pope who Ow the oe (AGE) and thee ar manor abuts evade. + The dae sl 6 eal ee However, bean we a gang 1 eit te ec inpr i gt ny fon the tla sy we | eset pytion rae sel ete +t ofl st ike what we with ny ter dat, wear gag port he Surat Hons datet and sr na vale called aa To import fm ‘shit sam we wl rad ran pe perenne eve anata) 4+ The Surat da for taney cota prfctor vr thatthe ample oe cn match apt hose ps, chee ul» peor hat can ipo ‘erate fae gn come rts of 1A hypothe 2 ay to doce» ing gros ta with dat The ype dines «perpen of ren X hat ou tet or vai Thao ished in both ene and machine eng nga. EB Pecorming te Hashing Tok +The class Fearetiher i highspeed owrmemory wear that wes « [higie oyna fae acing othe "hang + tte of ling 2 Hah ble of he fs enn In tng the veces ntar of Fentutaer pis hah ton the ets een er clr indocn sample mac Seely The ret i arena pod and eu ema ugh att apse of ‘eepecatiy te hate ons ot ee wht he np elo ke RRO nee