0
$\begingroup$

I am new and self-starter with the time series. How should I set my ARMA parameters with these ACF/PACF charts?enter image description here

EDIT: The actual data below:

2016-01-02,39.75951351903896
2016-01-03,44.19802892451088
2016-01-04,35.887366090361105
2016-01-05,47.52400693730769
2016-01-06,31.983702722150007
2016-01-07,18.057180708258063
2016-01-08,22.655207573981325
2016-01-13,24.032670139522725
2016-01-14,36.12322561430137
2016-01-15,20.187763237565214
2016-01-16,48.855922908363624
2016-01-17,48.39927291059999
2016-01-18,56.44049644907896
2016-01-19,37.69458044227272
2016-01-20,-70.50059842571429
2016-01-21,26.9161037942807
2016-01-22,43.75994638408695
2016-01-23,48.373229398888896
2016-01-24,12.572153745333333
2016-01-25,48.89474204682609
2016-01-26,28.538047387723083
2016-01-27,34.30517480442857
2016-01-28,47.78075214078947
2016-01-29,-78.68819873655555
2016-01-30,-28.97592228
2016-01-31,-7.046974465
2016-02-02,37.29205861404628
2016-02-03,42.25382787887097
2016-02-04,65.94607071840626
2016-02-05,12.651954309425003
2016-02-06,46.63573759473846
2016-02-07,32.25278198860416
2016-02-08,34.77058575470422
2016-02-14,-25.36701266
2016-02-15,-60.79984227216667
2016-02-16,56.02435255765384
2016-02-17,28.746407145797463
2016-02-18,16.550378951229508
2016-02-19,15.82792759401235
2016-02-20,4.990341175500004
2016-02-21,70.1803637223889
2016-02-22,31.350118170803576
2016-02-23,-86.2847876852143
2016-02-24,23.342789667022725
2016-02-25,39.77064584494442
2016-02-26,51.5940791500274
2016-02-27,51.88317634523076
2016-02-28,37.16568923229413
2016-02-29,-47.748961607533325
2016-03-02,21.166933922139705
2016-03-03,33.240946892873026
2016-03-04,46.14479273363332
2016-03-05,47.681106868636384
2016-03-06,34.2138656794186
2016-03-07,59.61628564819644
2016-03-08,50.47316287450001
2016-03-13,-65.29583746075001
2016-03-14,-34.110463839375
2016-03-15,49.83565921616666
2016-03-16,51.81790938512626
2016-03-17,32.48205597794505
2016-03-18,15.126982130323526
2016-03-19,-145.59020894999998
2016-03-20,-82.70413420999999
2016-03-21,30.991773486115378
2016-03-22,25.136369381434783
2016-03-23,9.599847528214283
2016-03-24,27.69982853081708
2016-03-25,-51.212730021000006
2016-03-26,-32.23208046
2016-03-28,-64.1687965
2016-03-29,45.859266805243244
2016-03-30,24.186714096475733
2016-03-31,34.76727254327955
2016-04-01,38.16134432345833
2016-04-02,35.63535579801471
2016-04-03,43.65374674151612
2016-04-04,17.67314425440425
2016-04-05,26.915555915125
2016-04-06,20.538772073199997
2016-04-07,20.102776686333335
2016-04-08,21.979722049833327
2016-04-13,45.928575165329676
2016-04-14,-32.48944289471428
2016-04-15,17.491105200894744
2016-04-16,64.69551535751515
2016-04-17,28.516430059499996
2016-04-18,40.95893484309175
2016-04-19,28.020746872985303
2016-04-20,30.130044503729412
2016-04-21,-58.18807158536842
2016-04-22,-25.337561036916664
2016-04-23,-26.627664854
2016-04-24,-28.69160864
2016-04-25,39.08488299530122
2016-04-26,46.435291118965225
2016-04-27,23.887642614118107
2016-04-28,-26.948563170055554
2016-04-29,19.046125541384615
2016-04-30,16.587165837604164
2016-05-01,13.155560922764712
2016-05-02,49.09833023170666
2016-05-03,40.75153917545162
2016-05-04,-11.1458260315
2016-05-05,20.513396592663636
2016-05-06,28.887168281440005
2016-05-07,37.936864492225
2016-05-08,33.74335998025926
2016-05-13,29.86262297474258
2016-05-14,38.24630225486047
2016-05-15,-40.245916527000006
2016-05-16,-41.557438196375
2016-05-17,24.794384094421687
2016-05-18,44.023555185375024
2016-05-19,5.268354276342105
2016-05-20,-40.386544623
2016-05-21,-72.32440148666667
2016-05-22,-65.92112163
2016-05-23,28.8193746211875
2016-05-24,50.234793735896915
2016-05-25,30.19691055108275
2016-05-26,19.686875528722226
2016-05-27,27.886102861170215
2016-05-28,15.590281383125
2016-05-29,39.651467377961545
2016-05-30,6.090026817734378
2016-05-31,25.362722162576084
2016-06-01,24.68601314195161
2016-06-02,36.179651624999984
2016-06-03,-26.168601790571433
2016-06-04,2.0638147895312495
2016-06-05,15.869253558152543
2016-06-06,30.97266805041892
2016-06-07,18.786122524463412
2016-06-08,62.50297165851221
2016-06-13,12.582850628634413
2016-06-14,17.844550084799994
2016-06-15,34.29525451988392
2016-06-16,19.03949167980612
2016-06-17,38.18312853862858
2016-06-18,-53.854408188923074
2016-06-19,-46.784044466875
2016-06-20,-29.811334852448272
2016-06-21,26.963462620788896
2016-06-22,33.756293176753616
2016-06-23,39.73173145911925
2016-06-24,50.64357838118183
2016-06-25,48.389665883354155
2016-06-26,9.094675531958334
2016-06-27,33.42214236674038
2016-06-28,25.31476515070093
2016-06-29,46.719154985418385
2016-06-30,30.091621642547025
2016-07-01,12.536105745310346
2016-07-02,42.674479050833334
2016-07-03,-13.53104158542857
2016-07-04,33.34080500357694
2016-07-06,33.21655168859616
2016-07-07,-87.41816937799999
2016-07-08,40.30393382614999
2016-07-13,34.82197758462265
2016-07-14,17.943902780436776
2016-07-15,25.110484135747257
2016-07-16,25.47542131667347
2016-07-17,26.70378401413636
2016-07-18,36.55949755945263
2016-07-19,21.239591675470592
2016-07-20,30.53004775844682
2016-07-21,38.65555894068421
2016-07-22,38.498780407582295
2016-07-23,64.67681735688889
2016-07-24,11.738560087312504
2016-07-25,-128.0038304
2016-07-26,37.739667621432446
2016-07-27,34.701622793094025
2016-07-28,-1.7783829946923093
2016-07-29,25.485521557819443
2016-07-30,50.257818642923084
2016-07-31,58.34552953883872
2016-08-01,31.165026671109075
2016-08-02,-22.0858203566111
2016-08-03,43.489689993647076
2016-08-04,31.459597491236853
2016-08-05,-5.974649260750001
2016-08-06,30.16309143433028
2016-08-07,25.860340325521754
2016-08-08,18.845143551831327
2016-08-13,27.317894660815796
2016-08-14,17.68989662983334
2016-08-15,33.30597511965824
2016-08-16,25.27378402822224
2016-08-17,19.28480775032501
2016-08-18,36.93733733110236
2016-08-19,39.781529773118805
2016-08-20,31.724542664115386
2016-08-21,60.21249229103573
2016-08-22,39.52576312805881
2016-08-23,27.766367975900902
2016-08-24,40.61433045760999
2016-08-25,48.67369527632799
2016-08-26,18.439678647561973
2016-08-27,30.46624090141937
2016-08-28,4.071043576733334
2016-08-29,24.64217842261111
2016-08-30,23.941792209643562
2016-08-31,19.871698113300756
$\endgroup$
  • 1
    $\begingroup$ Is that perhaps a homework exercise? If so, add the self-study and read its Wiki. Then tell us what you think about your problem so far, and we will help you advance. $\endgroup$ – Richard Hardy Dec 4 '16 at 17:50
  • $\begingroup$ It is a training exercise.. I have no lags out of boundaries.. so from my point of view there is no autocorrelation in this TS. Is that correct? $\endgroup$ – Mike_r_r Dec 4 '16 at 18:12
  • $\begingroup$ So far so good. $\endgroup$ – Richard Hardy Dec 4 '16 at 18:19
  • $\begingroup$ It's a time series on a daily purchase. I should obtain a out-of-sample forecast 14 days into future. Starting from those graphs I was thinking to use a ARMA(19,18)... But I think I am completely wrong.. $\endgroup$ – Mike_r_r Dec 4 '16 at 18:22
  • 1
    $\begingroup$ Yes, which means ARMA(p,q) with p=? and q=? $\endgroup$ – Richard Hardy Dec 4 '16 at 18:33
0
$\begingroup$

You have daily data . A cardinal rule when trying to form an ARIMA model is that you need equally spaced observations and no missing values . If the # of missing values are few then one can always fill in a zero and a replacement value can often be suggested (pulse anomaly). You have a number of missing days so all bets are off . Please fill in the missing days as best as you can as you might have day-to-day auto-correlation and potentially deterministic structure due to particular days-of-the-week etc . Please re-post your complete time series and your corrected acf and pacf.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.