Turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

- JMP User Community
- :
- Discussions
- :
- Standardization of the raw time series in JMP

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Standardization of the raw time series in JMP

Oct 28, 2018 6:32 PM
(1707 views)

Dear all,

I would like standardize my raw time series using a spline function of 32 years. How can I do in JMP?

Moreover, I would like to know how I could be able to obtain the EPS and Rbar statistical indeces. Generally, these indeces are commonly use in ARSTAN software for validate the standardization.

Regards,

Angelo

1 REPLY 1

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: Standardization of the raw time series in JMP

Created:
Oct 29, 2018 1:46 PM
| Last Modified: Oct 29, 2018 1:51 PM
(1669 views)
| Posted in reply to message from AngeloF88 10-28-2018

Angello,

A web search of ARSTAN software and EPS, lead me to this link ARSTAN_link which displays the output. From this and another discussion it appears ARSTAN is specialized software used to find relationships of environmental factors that might affect vegetation, growth or stress (tree rings) etc. This link ARSTAN_vs_R_dplr compares ARSTAN results to the R package dplr. It explains how they differ. The point is that both programs help organize the data from different sources to get trends so that relationships can be established.

It is likely that JMP can perform the analyses your need, but it will not be a turn-key application, unless someone has written a JMP application. Below, I have a script that might help, but will not be a perfect match. The environmental response data typically have multiple samples per time period or location. The example I have has just has one value per month, so the script creates a multiple sample per period year quarter, then uses a Gaussian (0.25) fit, saves the predicteds and residuals.

Since I am not an expert in this area, this is all I have. I hope it provides a direction to investigate. Note, if you have used dplr, you can use JMP to call up R routines.

```
Names Default to Here(1);
dt = Open("$sample_data\time series\steel shipments.jmp");
dt << New Column("YearQtr", numeric, Formula(Col Mean(:Date,Year(:Date), Quarter(:Date))),
Format("yyyyQq"));
biv = Bivariate(
Y( :Steel Shipments ),
X( :YearQtr ),
Fit Each Value( {Line Color( {66, 112, 221} )} ),
Kernel Smoother( 1, 4, 0.25, 0 )
);
biv << (curve[2] << Save Predicteds );
biv << (curve[2] << Save Residuals );
biv << (curve[2] << Line Color("Medium Dark Red") );
biv << (curve[2] << Line Width(3) );
gb = dt << Graph Builder(
Size( 534, 454 ),
Show Control Panel( 0 ),
Variables(
X( :YearQtr ),
Y( :Residuals Steel Shipments ),
Y( :Predicted Steel Shipments )
),
Elements(
Position( 1, 1 ),
Points(
X,
Y,
Legend( 14 ),
Summary Statistic( "Mean" ),
Error Bars( "Range" )
),
Line( X, Y, Legend( 15 ) )
),
Elements(
Position( 1, 2 ),
Points( X, Y, Legend( 9 ) ),
Line( X, Y, Legend( 12 ) )
),
SendToReport(
Dispatch(
{},
"400",
ScaleBox,
{Legend Model( 14, Base( 1, 0, 0, Item ID( "Range", 1 ) ) ),
Legend Model(
9,
Base( 0, 0, 0, Item ID( "Predicted Steel Shipments", 1 ) )
)}
),
Dispatch(
{},
"Graph Builder",
FrameBox,
{Grid Line Order( 2 ), Reference Line Order( 3 )}
),
Dispatch(
{},
"400",
LegendBox,
{Legend Position( {14, [-3, -1], 15, [2], 9, [0], 12, [1]} ),
Position( {-3, -1, 2, 0, 1} )}
)
)
);
```

Article Labels

There are no labels assigned to this post.