Time Series Practical Session Part- 3

SAS Analytics Time Series - Case Study & Practical Session
12 minutes
Share the link to this page
Copied
  Completed
You need to have access to the item to view this lesson.
One-time Fee
$69.99
List Price:  $99.99
You save:  $30
€64.60
List Price:  €92.30
You save:  €27.69
£55.40
List Price:  £79.14
You save:  £23.74
CA$95.14
List Price:  CA$135.92
You save:  CA$40.78
A$107.21
List Price:  A$153.17
You save:  A$45.95
S$94.29
List Price:  S$134.71
You save:  S$40.41
HK$547.55
List Price:  HK$782.26
You save:  HK$234.70
CHF 63.39
List Price:  CHF 90.56
You save:  CHF 27.17
NOK kr753.41
List Price:  NOK kr1,076.34
You save:  NOK kr322.93
DKK kr481.90
List Price:  DKK kr688.47
You save:  DKK kr206.56
NZ$116.54
List Price:  NZ$166.49
You save:  NZ$49.95
د.إ257.02
List Price:  د.إ367.19
You save:  د.إ110.16
৳7,699.83
List Price:  ৳11,000.23
You save:  ৳3,300.40
₹5,834.28
List Price:  ₹8,335.05
You save:  ₹2,500.76
RM331.26
List Price:  RM473.25
You save:  RM141.99
₦97,755.03
List Price:  ₦139,656.03
You save:  ₦41,901
₨19,509.87
List Price:  ₨27,872.43
You save:  ₨8,362.56
฿2,546.30
List Price:  ฿3,637.73
You save:  ฿1,091.43
₺2,255.86
List Price:  ₺3,222.80
You save:  ₺966.93
B$348.35
List Price:  B$497.67
You save:  B$149.31
R1,321.80
List Price:  R1,888.37
You save:  R566.56
Лв126.46
List Price:  Лв180.67
You save:  Лв54.20
₩94,455.35
List Price:  ₩134,942
You save:  ₩40,486.65
₪256.23
List Price:  ₪366.06
You save:  ₪109.83
₱3,937.46
List Price:  ₱5,625.18
You save:  ₱1,687.72
¥10,582.69
List Price:  ¥15,118.78
You save:  ¥4,536.09
MX$1,164.09
List Price:  MX$1,663.06
You save:  MX$498.96
QR255.79
List Price:  QR365.44
You save:  QR109.64
P961.07
List Price:  P1,373.02
You save:  P411.94
KSh9,203.68
List Price:  KSh13,148.68
You save:  KSh3,945
E£3,306.33
List Price:  E£4,723.53
You save:  E£1,417.20
ብር3,986
List Price:  ብር5,694.53
You save:  ብር1,708.53
Kz58,526.40
List Price:  Kz83,612.73
You save:  Kz25,086.33
CLP$68,613.99
List Price:  CLP$98,024.19
You save:  CLP$29,410.20
CN¥505.72
List Price:  CN¥722.49
You save:  CN¥216.77
RD$4,149.44
List Price:  RD$5,928.03
You save:  RD$1,778.58
DA9,428.56
List Price:  DA13,469.95
You save:  DA4,041.39
FJ$157.67
List Price:  FJ$225.25
You save:  FJ$67.58
Q547.27
List Price:  Q781.84
You save:  Q234.57
GY$14,690.75
List Price:  GY$20,987.69
You save:  GY$6,296.93
ISK kr9,671.91
List Price:  ISK kr13,817.61
You save:  ISK kr4,145.70
DH707.37
List Price:  DH1,010.57
You save:  DH303.20
L1,234.61
List Price:  L1,763.81
You save:  L529.19
ден3,973.74
List Price:  ден5,677.01
You save:  ден1,703.27
MOP$565.34
List Price:  MOP$807.66
You save:  MOP$242.32
N$1,327.13
List Price:  N$1,895.98
You save:  N$568.85
C$2,582.33
List Price:  C$3,689.21
You save:  C$1,106.87
रु9,350.11
List Price:  रु13,357.88
You save:  रु4,007.76
S/259.70
List Price:  S/371.02
You save:  S/111.31
K268.50
List Price:  K383.60
You save:  K115.09
SAR262.50
List Price:  SAR375.02
You save:  SAR112.51
ZK1,783.85
List Price:  ZK2,548.47
You save:  ZK764.61
L321.17
List Price:  L458.83
You save:  L137.66
Kč1,633.99
List Price:  Kč2,334.38
You save:  Kč700.38
Ft25,496.77
List Price:  Ft36,425.52
You save:  Ft10,928.74
SEK kr741.62
List Price:  SEK kr1,059.50
You save:  SEK kr317.88
ARS$59,963.93
List Price:  ARS$85,666.43
You save:  ARS$25,702.49
Bs486.54
List Price:  Bs695.09
You save:  Bs208.54
COP$272,760.10
List Price:  COP$389,673.99
You save:  COP$116,913.89
₡35,355.24
List Price:  ₡50,509.65
You save:  ₡15,154.41
L1,731.96
List Price:  L2,474.33
You save:  L742.37
₲515,624.88
List Price:  ₲736,638.54
You save:  ₲221,013.66
$U2,646.58
List Price:  $U3,781
You save:  $U1,134.41
zł278.71
List Price:  zł398.17
You save:  zł119.46
Already have an account? Log In

Transcript

We have started with the practical session of time series analysis in SAS software we were dealing with the SAS data set that is ln data set which consists of two variables data and passengers. It consists of the number of passengers that it contains the number of passengers that has traveled from the month 1949 that is from the date Jan 1949 till December 1960, that is it contains the number of passengers that has traveled through Jet Airways Ellen company from the month Jan 2014, till the month December 1960. And our objective is to predict for the coming 12 months So, for that we are going to try modeling to at first we have checked stationarity of the data through scatterplot. Then we had done Arima modeling we saw the cnpc graphs, the We did a DNA test that is augmented Dickey fuller test to check non stationarity all the three tests shows that the data is very much non stationary in nature, then we then we did animal modeling with different periods of differencing.

First we did animal modeling with period of differencing one then period two and then Piero differencing as one comma one in order to gradually move towards stationarity. We have seen that our ACF graphs a PC graphs were gradually changing is verging gradually moving towards stationarity we also found a huge plunge in autocorrelation between one and 12 lakh in our first resolver which we had got through animal modeling. That is this one. See there is a huge plunge in autocorrelation in the first 1112 So, now we are going to use this one comment with as the accuracy of differencing to store the stationarity. So, let's do that. We are going to use the procedure proc Arima data equals my new corn time series underscore line identify VOD passengers.

I'm using the pH DS one comma 12. That is a appropriate period of difference in which we had caught enough first result viewer see weird Then run and then quit let's run this code see her obviously visa process completely changed the lens of autocorrelation length of come between the within the standard error region that means we have distraught stationarity that is our data has distribution at this is our autocorrelation table. Now the behavioral differences comment with no working CDC 0.13206 standard deviations with countries or an infinite number of observations from 30 that is at one comma 12 behavioral difference in 13 observations are elevated. So, the meaning observations are from 31. Now, we are going to do the mini stable Just because we have restored stationarity now we'll be using the mini table to find the optimal answer for a model and a model. So we'll be using the same procedure proc Arima later.

Equals like rename is mainly been done dark data set in a nice time series underscore and name var equal to passengers passengers is man and become available. I'm reading that period of difference he has one comment well then I'm using the Keyboard mini mini stands for minimum information criteria. So we'll be using the mini table to find optimal Elsevier and a model from and then we could execute the code Let me explain on the code. I'm using the procedure Procurement Data on the library on the data set entities underscore ln which is located inside the library mainly my analytical variable passengers. So use the statement I will say var equal to passengers one comment when we set up a period or difference in that we have got in our last reserved viewer, that is towards the beginning. This is huge planter autocorrelation at first when it was 0.9 coral over here and zero to 60 days a huge plant so we'll be using one common terms that refer to differencing.

Then we are using the key word meaning that stands for minimum information criteria. Many people is used to document lakhs of arnn a model will be using this table and then run a little bit. So now let's execute this code. See this is many, many people available information criteria table here are the minimum table values given these values are basically pac, pac stands for Bayesian information criteria, which explains the amount of unexplained variation in our model. So, lesser is the PAC value better is a model because lesser is unexplained variation better is the model over here the minimum Pac values given us hope on a 369. So, this is the minimum basic minimum value of all the values in the table.

So optimal level VR will be one and optimal level a will be zero. So optimal level VR model that is P, P will be equal to one and Q will be zero. So P and Q are lsv got this optimal level v nav model This is the optimal relationship for like no stimulus given period of differences. One comment with me no CDs, same 0.13206, which was before it, I'm a modeling with one common difference externality deviation. Again, it's the same number of observations to the same one that is 13. observations were eliminated by doing the differencing. And one common trend.

So our optimal lots is areas one p equal to one, if you will, to see those that we have to move down. So let's note it down because we have to use that for forecasting the variables to one and q1 is equal to zero. That's notice Don't be using it. We'll be forecasting by applying this optimal less desirable disarmament Using the same procedure property manager will be focusing for the coming 12 months my lab reading is against in my live one.com citizen discord bar as he just as he uses meat and unethical variables that are being used very very imagery blasting and general referencing which is document one that is one comment well then I will give document labs estimate the values as I told as we calculated is equal to one from the meaning table and Q value is equal to Zero via document every animal case or lack of a model.

We haven't forecast for the coming 12 months from reading forecast lead equals to 12. Thank you. So, let me explain the code. Here we are doing Arima modeling on the data set MCs underscore nn located in segment one, we are using passages variable which is an analytical variable our auto references can comment where there is a lot of fencing, which we got in our beginning in the first result viewer where the plunging on conditions between one comma 12 So, this was the planche we obtained in the difference in the span command Well, from the mini table we got the optimum lots of air and in a model that is from the minimum be AC very open a three string air model is run an animal to zero that is equal to one to zero. And we are forecasting for the coming 12 months to forecast legal structure.

So let's run this code should be a semicolon over here. No Let's run the score. To see this is a forecast from video. passages in our original data set we had till 144 observation that is from 2014 jan jan 1842, December 2016, Mr. Cooperman 44 observations per 141 words it was dis forecasted. So, it is forecasted from 45 onwards these are these are the forecast values for the coming 12 months this is standard error This is an indicator percent confidence limits, we got the autoregressive factors because the estimated mean the period or differencing is one comment well these are the graphs these are the seven basic graphs. This is the forecast for passengers observations with the forecast oils now, we'll be learning to hear in this video.

And with this video we have completed the scripted modeling program. I hope you all have enjoyed this session. Thank you Goodbye and All the best to you

Sign Up

Share

Share with friends, get 20% off
Invite your friends to LearnDesk learning marketplace. For each purchase they make, you get 20% off (upto $10) on your next purchase.