summary.sisal: Summarizing Sequential Input Selection Results
In sisal: Sequential Input Selection Algorithm

Description Usage Arguments Details Value Author(s) See Also Examples

summary method for class "sisal"

## S3 method for class 'sisal'
summary(object, ...)
## S3 method for class 'summary.sisal'
print(x, ...)

`object`	an object of class `"sisal"`.
`x`	an object of class `"summary.sisal"`.
`...`	arguments passed to/from other methods.

The functions compute and print summaries (summary.lm) of the ordinary least squares regression models stored in the object and some additional information.

The function summary.sisal returns a list with class "summary.sisal", currently containing:

`summ.full`	summary of the full model. An object of class `"summary.lm"`.
`summ.L.v`	summary of the `L.v` model. An object of class `"summary.lm"`.
`summ.L.f`	summary of the `L.f` model. An object of class `"summary.lm"`.
`error.df`	a `data.frame` containing information on the best variable sets with a given number of variables, with the following columns (copied from `object`): n.inputs number of inputs (row label). E.tr mean training MSE. s.tr standard deviation of training MSE. E.v mean validation MSE. L.f.flag `logical` `vector` where the location of `TRUE` points the smallest variable set with `thr.flag` `TRUE`. L.v.flag `logical` `vector` where the location of `TRUE` points the variable set with the smallest validation error. thr.flag `logical` `vector` where `TRUE` means that error is at most `E.v[L.v.flag] + s.tr[L.v.flag]`.

The function print.summary.sisal invisibly returns x.

Mikko Korpela

sisal, print.sisal

1 2	foo <- testSisal(dataset="toy", Mtimes=10, hbranches=2) summary(foo)

Dataset is "toy"
See ?toy.learn for a description of the data
Adding noise with standard deviation 0.200000
Calling sisal

pruning.criterion=round robin, Mtimes=10, kfold=10, hbranches=2, max.width=4, 
q=0.165, standardize=TRUE, pruning.keep.best=TRUE, pruning.reverse=FALSE, 
use.ridge=FALSE, sp=-1
There are 10 variables and 1000 samples.
There are no missing values.
================================================================================summary() for linear models using all samples
=============================================
All inputs
----------

Call:
lm(formula = lmForm, data = as.data.frame(X.complete), na.action = na.exclude)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.57609 -0.13943 -0.00138  0.13834  0.61647 

Coefficients:
              Estimate Std. Error t value Pr(>|t|)    
(Intercept)  8.069e-19  6.225e-03   0.000   1.0000    
X1           2.618e-01  6.264e-03  41.800   <2e-16 ***
X2           5.361e-01  6.236e-03  85.961   <2e-16 ***
X3           7.840e-01  6.256e-03 125.312   <2e-16 ***
X4           1.235e-03  6.247e-03   0.198   0.8434    
X5           6.917e-03  6.276e-03   1.102   0.2707    
X6          -1.007e-04  6.264e-03  -0.016   0.9872    
X7          -5.348e-03  6.261e-03  -0.854   0.3932    
X8          -9.944e-03  6.252e-03  -1.591   0.1120    
X9          -4.074e-03  6.261e-03  -0.651   0.5154    
X10         -1.171e-02  6.264e-03  -1.869   0.0619 .  
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 0.1969 on 989 degrees of freedom
Multiple R-squared:  0.9616,	Adjusted R-squared:  0.9612 
F-statistic:  2479 on 10 and 989 DF,  p-value: < 2.2e-16

L.v inputs
----------

Call:
lm(formula = lmForm, data = as.data.frame(X.L.v), na.action = na.exclude)

Residuals:
     Min       1Q   Median       3Q      Max 
-0.60331 -0.14220  0.00009  0.14071  0.59777 

Coefficients:
              Estimate Std. Error t value Pr(>|t|)    
(Intercept)  6.168e-19  6.222e-03   0.000   1.0000    
X1           2.628e-01  6.230e-03  42.188   <2e-16 ***
X2           5.361e-01  6.227e-03  86.103   <2e-16 ***
X3           7.846e-01  6.228e-03 125.985   <2e-16 ***
X10         -1.246e-02  6.232e-03  -1.999   0.0459 *  
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 0.1968 on 995 degrees of freedom
Multiple R-squared:  0.9614,	Adjusted R-squared:  0.9613 
F-statistic:  6202 on 4 and 995 DF,  p-value: < 2.2e-16

L.f inputs
----------

Call:
lm(formula = lmForm, data = as.data.frame(X.L.f), na.action = na.exclude)

Residuals:
    Min      1Q  Median      3Q     Max 
-0.5947 -0.1408  0.0001  0.1422  0.5992 

Coefficients:
             Estimate Std. Error t value Pr(>|t|)    
(Intercept) 4.414e-19  6.231e-03    0.00        1    
X1          2.624e-01  6.235e-03   42.08   <2e-16 ***
X2          5.359e-01  6.235e-03   85.95   <2e-16 ***
X3          7.843e-01  6.235e-03  125.78   <2e-16 ***
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 0.1971 on 996 degrees of freedom
Multiple R-squared:  0.9613,	Adjusted R-squared:  0.9612 
F-statistic:  8244 on 3 and 996 DF,  p-value: < 2.2e-16

Errors, measured with cross-validation
======================================
 #Inputs  Tr. error       Tr. SD Val. error   
       0 0.99890346 0.0150527856 1.00083421   
       1 0.39318305 0.0061534877 0.39490956   
       2 0.10739338 0.0015390797 0.10818198   
       3 0.03865907 0.0005376807 0.03897912 **
       4 0.03850082 0.0005304519 0.03889167 **
       5 0.03839847 0.0005345437 0.03889541  *
       6 0.03836668 0.0005335045 0.03893341  *
       7 0.03831352 0.0005312059 0.03898488  *
       8 0.03829318 0.0005326125 0.03904502  *
       9 0.03828758 0.0005315167 0.03912195  *
      10 0.03828234 0.0005312983 0.03922295  *
Code: one "*" for each criterion:
* error is inside threshold
* smallest validation error (L.v)
* least complex model with error inside threshold (L.f)