Multivariate Methods • Correlations and Multivariate Techniques • Launch the Multivariate Platform
Previous
 • 
Next
Launch the Multivariate Platform
Launch the Multivariate platform by selecting Analyze > Multivariate Methods > Multivariate.
Figure 2.2 The Multivariate Launch Window
Y, Columns
Defines one or more response columns.
Weight
Identifies one column whose numeric values assign a weight to each row in the analysis.
Freq
Identifies one column whose numeric values assign a frequency to each row in the analysis.
By
Performs a separate multivariate analysis for each level of the By variable.
Estimation Method
Specifies the method for calculating the correlations. REML and Pairwise are the methods used most frequently. Several of these methods address the treatment of missing data. You can also estimate missing values by using the estimated covariance matrix, and then using the Impute Missing Data command. See Impute Missing Data.
Default
The Default option uses either the Row-wise, Pairwise, or REML methods.
–
Row-wise estimation is used for data tables with no missing values.
–
Pairwise estimation is used for data tables with missing values and either more than 10 columns, more than 5,000 rows, or more columns than rows.
–
REML estimation is used otherwise.
Note: When the Default option would otherwise result in REML estimation, but the fit does not converge properly, the platform reverts to the Pairwise method. This can happen when there are missing values in your data table and at least one of the following situations applies: if your data table has fewer than 10 columns, fewer than 5,000 rows, or fewer columns than rows. If the estimation method shown is Pairwise, this means that the REML fit did not converge.
REML
Restricted maximum likelihood (REML) estimation uses all of the data, even if missing values are present. Due to a bias-correction factor, this method is slow if the dataset is large and there are many missing values. Therefore, REML is most useful for smaller datasets. If there are no missing cells in the data, then the REML and ML estimates are equivalent and equal to the sample covariance matrix. If there are missing cells, REML’s variance and covariance estimates are less biased than the estimates from ML estimation. For more information, see REML.
ML
Maximum likelihood (ML) estimation uses all of the data, even if missing values are present. Because the estimates from ML are generated quickly, this method is most useful for large data tables with missing data.
Robust
Robust estimation uses all of the data, even if missing values are present. This method down-weights extreme values and is therefore useful for data tables that might have outliers. For statistical details, see Robust.
Row-wise
Row-wise estimation does not use observations with missing values, so rows that contain missing cells are deleted before the method is applied. This method is useful for excluding any observations that have missing data. Row-wise estimation was the only estimation method available prior to JMP 8, so it can also be used to check compatibility with JMP versions prior to JMP 8.
Pair-wise
Pair-wise estimation uses all of the data, even if missing values are present. This estimation method performs correlations for all rows for each pair of columns with nonmissing values. It is most useful when a data table has missing values and either more columns than rows, more than 10 columns, or more than 5,000 rows.
Note: If you select REML, ML, or Robust and your data table contains more columns than rows and has missing values, JMP switches the Estimation Method to Pairwise.
Matrix Format
Select a format option for the Scatterplot Matrix. The Square option displays plots for all ordered combinations of columns. Lower Triangular displays plots below the diagonal, with the first n - 1 columns on the horizontal axis. Upper Triangular displays plots above the diagonal, with the first n - 1 columns on the vertical axis.

Previous
 • 
Next
Help created on 7/12/2018