Type Members

1.  case class Glm(y: DenseVector[Double], Xmat: DenseMatrix[Double], colNames: Seq[String], fam: GlmFamily, addIntercept: Boolean = true, its: Int = 50) extends Model with Product with Serializable

   Generalised linear regression modelling

   Generalised linear regression modelling

   y

   Vector of responses

   Xmat

   Covariate matrix

   colNames

   List of covariate names

   fam

   Observation model. eg. LogisticGlm or PoissonGlm

   addIntercept

   Add an intercept term to the covariate matrix?

   its

   Max iterations for the IRLS algorithm (default 50)

   returns

   An object of type Glm with many useful methods providing information about the regression fit, including .coefficients, .p and .summary

2.  sealed trait GlmFamily extends AnyRef

   Trait for simple one-parameter exponential family observation models.

3.  case class Lm(y: DenseVector[Double], Xmat: DenseMatrix[Double], colNames: Seq[String], addIntercept: Boolean = true) extends Model with Product with Serializable

   Linear regression modelling

   Linear regression modelling

   y

   Vector of responses

   Xmat

   Covariate matrix

   colNames

   List of covariate names

   addIntercept

   Add an intercept term to the covariate matrix?

   returns

   An object of type Lm with many useful attributes providing information about the regression fit

4.  trait Model extends AnyRef
5.  case class Pca(mat: DenseMatrix[Double], colNames: Seq[String]) extends Product with Serializable

   Principal components analysis

   Principal components analysis

   Computed using SVD of the centred data matrix rather than from the spectral decomposition of the covariance matrix. eg. More like the R function "prcomp" than the R function "princomp".

   NOTE: .loadings are transposed relative to the PCA function in Breeze

   mat

   Data matrix with rows corresponding to observations and columns corresponding to variables

   colNames

   Sequence of column names of mat

   returns

   An object of type Pca with methods such as .loadings, .scores, .sdev and .summary

6.  trait Predict extends AnyRef
7.  case class PredictGlm(mod: Glm, newX: DenseMatrix[Double], response: Boolean) extends Predict with Product with Serializable

   Prediction from a fitted linear (Glm) model

   Prediction from a fitted linear (Glm) model

   mod

   fitted generalised linear model

   newX

   covariate matrix for predictions

   response

   predictions on the response scale?

   returns

   An object of type PredictGlm with several useful attributes, including .fitted and .se

8.  case class PredictLm(mod: Lm, newX: DenseMatrix[Double]) extends Predict with Product with Serializable

   Prediction from a fitted linear (Lm) model

   Prediction from a fitted linear (Lm) model

   mod

   fitted linear model

   newX

   covariate matrix for predictions

   returns

   An object of type PredictLm with several useful attributes, including .fitted and .se