The folda
package is an R modeling tool designed for
fitting Forward Stepwise Linear Discriminant Analysis (LDA) and
Uncorrelated Linear Discriminant Analysis (ULDA). If you’re unfamiliar
with stepwise LDA or ULDA, please refer to the following resources:
For stepwise LDA using Wilks’ Lambda, see Section 6.11.1 in Methods of Multivariate Analysis, Third Edition by Alvin C. Rencher and William F. Christensen (2012).
For ULDA, refer to Ye, J., & Yu, B. (2005). Characterization of a family of algorithms for generalized discriminant analysis on undersampled problems. Journal of Machine Learning Research, 6(4). Link.
For a combination of ULDA and forward LDA using Pillai’s trace, see Wang, S. (2024). A New Forward Discriminant Analysis Framework Based on Pillai’s Trace and ULDA. arXiv preprint arXiv:2409.03136. Link.
folda
package?If you’ve ever been frustrated by the warnings and errors from
MASS::lda()
, you will appreciate the ULDA implementation in
folda()
. It offers several key improvements:
No more “constant within group” errors! ULDA can handle constant columns and perfect separation of groups.
Automatic missing value handling! The implementation seamlessly integrates automatic missing value imputation during both training and testing phases.
Fast! ULDA is implemented using the Generalized
Singular Value Decomposition (GSVD) method, which diagonalizes both
within-class and total scatter matrices simultaneously, offering a speed
advantage over the sequential diagonalization used in
MASS::lda()
(see Howland et al., 2003 for more details). We
have also rewritten the matrix decomposition modules (SVD, QR) using
RcppEigen
, further improving computational efficiency by
leveraging optimized C++ code.
Better visualization! folda
uses
ggplot2
to provide visualizations of class separation in
projected 2D spaces (or 1D histograms), offering valuable
insights.
For the forward LDA implementation, folda
offers the
following advantages over the classical framework:
No issues with multicollinearity or perfect linear
dependency! Since folda()
is built on ULDA, it
effectively solves for the scaling matrix.
Handles perfect separation and offers greater
power! The classical approach using Wilks’ Lambda has known
limitations, including premature stopping when some (not all) groups are
perfectly separated. Pillai’s trace, as used in folda()
,
not only effectively addresses perfect separation, but has also been
shown to generally have greater statistical power than Wilks’ Lambda
(Rencher et al., 2002).
folda
library(folda)
mpg <- as.data.frame(ggplot2::mpg) # Prepare the data
datX <- mpg[, -5] # All predictors without Y
response <- mpg[, 5] # we try to predict "cyl" (number of cylinders)
Build a ULDA model with all variables:
Build a ULDA model with forward selection via Pillai’s trace:
fit <- folda(datX = datX, response = response, subsetMethod = "forward", testStat = "Pillai")
print(fit) # 6 out of 11 variables are selected, displ is the most important among them
#>
#> Overall Pillai's trace: 1.325
#> Associated p-value: 4.636e-74
#>
#> Prediction Results on Training Data:
#> Refitting Accuracy: 0.9188
#> Gini Index: 0.7004
#>
#> Confusion Matrix:
#> Actual
#> Predicted 4 5 6 8
#> 4 69 0 3 0
#> 5 8 4 2 0
#> 6 4 0 74 2
#> 8 0 0 0 68
#>
#> Group means of LD scores:
#> LD1 LD2 LD3
#> 4 3.05298379 0.02700248 -0.3555829
#> 5 1.87744449 -4.45014946 0.8156167
#> 6 0.06757888 0.28356907 0.5911862
#> 8 -3.71628852 -0.09697943 -0.3023424
#>
#> Forward Selection Results:
#> var statOverall statDiff threshold
#> 1 displ 0.873393 0.87339300 0.06545381
#> 2 modelnew beetle 1.029931 0.15653777 0.05673510
#> 3 modeljetta 1.141651 0.11172064 0.05496185
#> 4 modelcaravan 2wd 1.210165 0.06851331 0.05363507
#> 5 classmidsize 1.263449 0.05328468 0.05276500
#> 6 cty 1.325255 0.06180560 0.05194279
Plot the results:
One-dimensional plot:
# A 1D plot is created when there is only one feature
# or for binary classification problems.
mpgSmall <- mpg[, c("cyl", "displ")]
fitSmall <- folda(mpgSmall[, -1, drop = FALSE], mpgSmall[, 1])
plot(fitSmall, mpgSmall, mpgSmall[, 1])
Make predictions:
head(predict(fit, datX, type = "response"))
#> [1] "4" "4" "4" "4" "6" "4"
head(predict(fit, datX, type = "prob")) # Posterior probabilities
#> 4 5 6 8
#> 1 0.9966769 7.475058e-08 0.0033230408 7.023764e-12
#> 2 0.9994438 1.401133e-08 0.0005562131 5.338710e-13
#> 3 0.9970911 3.835722e-08 0.0029088506 1.738154e-11
#> 4 0.9983963 2.196016e-08 0.0016037009 7.365641e-12
#> 5 0.3122116 6.809673e-07 0.6877815595 6.173116e-06
#> 6 0.5995781 4.275271e-07 0.4004193019 2.123291e-06
Here, we compare their performances in a scenario where one group of classes is perfectly separable from another, a condition under which Wilks’ Lambda performs poorly. Now let’s predict the model of the car.
fitW <- folda(mpg[, -2], mpg[, 2], testStat = "Wilks")
fitW$forwardInfo
#> var statOverall statDiff threshold
#> 1 manufactureraudi 0 0 0.7338759
Wilks’ Lambda only selects manufacturer-audi, since it can separate a4, a4 quattro, and a6 quattro from other models. However, it unexpectedly stops since the Wilks’ Lambda = 0, leading to a refitting accuracy of 0.0812.
fitP <- folda(mpg[, -2], mpg[, 2], testStat = "Pillai")
fitP$forwardInfo
#> var statOverall statDiff threshold
#> 1 manufacturerchevrolet 1.00000 1.0000000 0.2654136
#> 2 manufacturerpontiac 2.00000 1.0000000 0.2599086
#> 3 classpickup 3.00000 1.0000000 0.2543730
#> 4 manufacturerdodge 4.00000 1.0000000 0.2488055
#> 5 manufacturerford 5.00000 1.0000000 0.2432051
#> 6 manufacturerhonda 6.00000 1.0000000 0.2375706
#> 7 manufacturerhyundai 7.00000 1.0000000 0.2319008
#> 8 manufacturerjeep 8.00000 1.0000000 0.2261943
#> 9 manufactureraudi 9.00000 1.0000000 0.2204496
#> 10 manufacturerland rover 10.00000 1.0000000 0.2146652
#> 11 manufacturerlincoln 11.00000 1.0000000 0.2088394
#> 12 manufacturermercury 12.00000 1.0000000 0.2029702
#> 13 manufacturernissan 13.00000 1.0000000 0.1970556
#> 14 manufacturersubaru 14.00000 1.0000000 0.1910933
#> 15 manufacturertoyota 15.00000 1.0000000 0.1850810
#> 16 drvf 16.00000 1.0000000 0.1790159
#> 17 drvr 17.00000 1.0000000 0.1723677
#> 18 classsuv 18.00000 1.0000000 0.1661696
#> 19 classminivan 19.00000 1.0000000 0.1599071
#> 20 classmidsize 19.93159 0.9315947 0.1535761
#> 21 classsubcompact 20.74392 0.8123229 0.1475693
#> 22 classcompact 21.71027 0.9663480 0.1421899
#> 23 displ 21.96954 0.2592742 0.1358348
#> 24 transauto(s5) 22.16831 0.1987676 0.1335988
#> 25 cty 22.35530 0.1869879 0.1316772
#> 26 flp 22.52155 0.1662562 0.1297761
On the other hand, Pillai’s trace selects 26 variables in total and
the refitting accuracy is 0.9231. Additionally, MASS::lda()
would throw an error in this scenario due to the “constant within
groups” issue.
The default method to handle missing values are
c(medianFlag, newLevel)
. It means that for numerical
variables, missing values are imputed with the median, while for
categorical variables, a new level is assigned to represent missing
values. Additionally, for numerical variables, we generate missing value
indicators to flag which observations had missing data.
Two key functions involved in this process are
missingFix()
and getDataInShape()
:
missingFix()
imputes missing values
and outputs two objects: the imputed dataset and a missing reference,
which can be used for future imputations. Any constant columns remains
in the imputed dataset will be removed.
getDataInShape()
takes new data and
the missing reference as inputs, and returns an imputed dataset. This
function performs several tasks:
# Create a dataset with missing values
(datNA <- data.frame(X1 = rep(NA, 5), # All values are NA
X2 = factor(rep(NA, 5), levels = LETTERS[1:3]), # Factor with all NA values
X3 = 1:5, # Numeric column with no missing values
X4 = LETTERS[1:5], # Character column
X5 = c(NA, 2, 3, 10, NA), # Numeric column with missing values
X6 = factor(c("A", NA, NA, "B", "B"), levels = LETTERS[1:3]))) # Factor with missing values
#> X1 X2 X3 X4 X5 X6
#> 1 NA <NA> 1 A NA A
#> 2 NA <NA> 2 B 2 <NA>
#> 3 NA <NA> 3 C 3 <NA>
#> 4 NA <NA> 4 D 10 B
#> 5 NA <NA> 5 E NA B
Impute missing values and create a missing reference:
(imputedSummary <- missingFix(datNA))
#> $data
#> X3 X4 X5 X6 X5_FLAG
#> 1 1 A 3 A 1
#> 2 2 B 2 new0_0Level 0
#> 3 3 C 3 new0_0Level 0
#> 4 4 D 10 B 0
#> 5 5 E 3 B 1
#>
#> $ref
#> X3 X4 X5 X6 X5_FLAG
#> 1 3 A 3 new0_0Level 1
X1 and X2 are removed because they are constant (i.e., all values are NA). X3 and X4 remain unchanged. X5 is imputed with the median (3), and a new column X5_FLAG is added to indicate missing values. X6 is imputed with a new level ‘new0_0Level’.
Now, let’s create a new dataset for imputation.
(datNAnew <- data.frame(X1 = 1:3, # New column not in the reference
X3 = 1:3, # Matching column with no NAs
X4 = as.factor(c("E", "F", NA)), # Factor with a new level "F" and missing values
X5 = c(NA, 2, 3))) # Numeric column with a missing value
#> X1 X3 X4 X5
#> 1 1 1 E NA
#> 2 2 2 F 2
#> 3 3 3 <NA> 3
Apply the missing reference to the new dataset:
getDataInShape(datNAnew, imputedSummary$ref)
#> X3 X4 X5 X6 X5_FLAG
#> 1 1 E 3 new0_0Level 1
#> 2 2 A 2 new0_0Level 0
#> 3 3 A 3 new0_0Level 0
X1 is removed because it does not exist in the missing reference. X3 remains unchanged. “F” is a new level in X4, so it is removed and imputed with “A” (the most frequent level) along with other missing values. X5 is imputed, and a new column X5_FLAG is added to indicate missing values. X6 is missing from the new data, so it is initialized with the level “new0_0Level”.
Next, we show an example using folda
with the airquality
dataset. First, let’s check which columns in airquality have missing
values:
sapply(airquality, anyNA) # Ozone and Solar.R have NAs
#> Ozone Solar.R Wind Temp Month Day
#> TRUE TRUE FALSE FALSE FALSE FALSE
Our response variable is the 5th column (Month):
The generated missing reference is:
fitAir$misReference
#> Ozone Solar.R Wind Temp Day Ozone_FLAG Solar.R_FLAG
#> 1 31.5 205 9.7 79 16 1 1
To make prediction:
Notice that no issues arise during predicting, even when the new data contains nothing but missing values.
There are two common scenarios where downsampling can be helpful:
The classes are highly imbalanced, and downsampling can make them more balanced.
To speed up computation. The ULDA classifier is computed based on class centroids and the covariance structure. Once a sufficient number of data points, such as 3000, are included, additional data points may have minimal impact.
By default, downsampling
is disabled. If
downsampling = TRUE
and kSample = NULL
, it
will downsample all classes to the size of the smallest class. If
kSample
is specified, all classes will be downsampled to
have a maximum of kSample
samples. An if a class contains
fewer than kSample
samples, all observations from that
class will be retained.
Suppose we want to predict the number of cylinders (cyl
)
in a car. The number of observations in each class is:
If we apply downsampling without specifying kSample
, we
will randomly select 4 samples from each group, as the smallest group
(cyl = 5
) has only 4 observations.
set.seed(443)
fitCyl <- folda(mpg[, -5], mpg[, 5], downSampling = TRUE)
fitCyl$confusionMatrix
#> Actual
#> Predicted 4 5 6 8
#> 4 4 0 0 0
#> 5 0 4 0 0
#> 6 0 0 4 0
#> 8 0 0 0 4
We can also set kSample = 30
. In this case, 30 random
samples will be selected from cyl = 4, 6, 8
, while 4
samples will be chosen from cyl = 5
.
fitCyl30 <- folda(mpg[, -5], mpg[, 5], downSampling = TRUE, kSample = 30)
fitCyl30$confusionMatrix
#> Actual
#> Predicted 4 5 6 8
#> 4 22 0 3 0
#> 5 8 4 2 0
#> 6 0 0 25 2
#> 8 0 0 0 28
It’s important to note that this downsampling process changes the prior, and all subsequent results are based on the downsampled data. If you only want to downsample for speed (or another reason while maintaining the original proportion or prior), be sure to specify the prior explicitly.
fitCylWithPrior <- folda(mpg[, -5], mpg[, 5], downSampling = TRUE, prior = table(mpg[, 5]))
fitCylWithPrior$confusionMatrix
#> Actual
#> Predicted 4 5 6 8
#> 4 4 4 1 0
#> 5 0 0 0 0
#> 6 0 0 3 0
#> 8 0 0 0 4
As we can see, the prior in this model suppresses the prediction of
class cyl = 5
, and the confusion matrix differs from the
one in fitCyl
.
correction
: If you’re less
concerned about controlling the type I error and prefer a more
aggressive variable selection process, setting
correction = FALSE
may result in better testing accuracy,
particularly when the number of columns exceeds the number of
rows.
alpha
: If your goal is to rank all
variables, set alpha = 1
. This ensures that no variables
are filtered out during the selection process.
misClassCost
: This parameter is
useful in situations where misclassifying certain classes has a more
severe impact compared to others. Below is an example demonstrating how
to incorporate different misclassification costs.
The iris dataset is a famous dataset with three species of flowers:
Suppose misclassifying versicolor into other species is very costly. A potential misclassification cost matrix might look like this:
This means that misclassifying versicolor to other species is 100 times more severe than misclassifying other species to versicolor. First, let’s fit the model with equal misclassification costs and specified misclassification costs:
fitEqualCost <- folda(iris[, -5], response = iris[, 5])
fitNewCost <- folda(iris[, -5], response = iris[, 5], misClassCost = misClassCost)
The prediction distributions with equal misclassification costs:
The prediction distributions with specified misclassification costs:
As shown, the model tends to predict versicolor more often due to the higher misclassification cost associated with predicting it incorrectly.
Howland, P., Jeon, M., & Park, H. (2003). Structure preserving dimension reduction for clustered text data based on the generalized singular value decomposition. SIAM Journal on Matrix Analysis and Applications, 25(1), 165-179.
Rencher, A. C., & Christensen, W. F. (2002). Methods of Multivariate Analysis (Vol. 727). John Wiley & Sons.
Wang, S. (2024). A new forward discriminant analysis framework based on Pillai’s trace and ULDA. arXiv preprint, arXiv:2409.03136. Retrieved from https://arxiv.org/abs/2409.03136.