scalation
API
Generated with
scalation/scalation/scalation.modeling/scalation.modeling.forecasting

scalation.modeling.forecasting

package scalation.modeling.forecasting

Members list

Packages

package scalation.modeling.forecasting.neuralforecasting

Type members

Classlikes

class AR(y: VectorD, tt: VectorD, hparam: HyperParameter) extends FitI, Forecaster, Correlogram

The AR class provides basic time series analysis capabilities for Auto-Regressive (AR) models. In an AR(p) model, p refers to the order of the Auto-Regressive components of the model. AR models are often used for forecasting. Given time series data stored in vector y, its next value y_t+1 = y(t+1) may be predicted based on prior values of y and its noise: y_t+1 = δ + Σ(φ_j y_t-j) + e_t+1 where δ is a constant, φ is the auto-regressive coefficient vector, and e_t+1 is the noise term.

The AR class provides basic time series analysis capabilities for Auto-Regressive (AR) models. In an AR(p) model, p refers to the order of the Auto-Regressive components of the model. AR models are often used for forecasting. Given time series data stored in vector y, its next value y_t+1 = y(t+1) may be predicted based on prior values of y and its noise: y_t+1 = δ + Σ(φ_j y_t-j) + e_t+1 where δ is a constant, φ is the auto-regressive coefficient vector, and e_t+1 is the noise term.

Value parameters

hparam

the hyper-parameters

tt

the time vector, if relevant (time index may suffice)

y

the response vector (time series data)

Attributes

Companion
object
Supertypes
trait Correlogram
trait Forecaster
trait Model
class FitI
trait Fit
trait FitM
class Object
trait Matchable
class Any
Show all
object AR

The AR companion object provides factory methods for the AR class. Use SARIMAX for hyper-parameters.

The AR companion object provides factory methods for the AR class. Use SARIMAX for hyper-parameters.

Attributes

Companion
class
Supertypes
class Object
trait Matchable
class Any
Self type
AR.type
class AR1MA(y: VectorD, tt: VectorD, hparam: HyperParameter, diffr: Boolean) extends Forecaster, Correlogram, Fit

The AR1MA class provides basic time-series analysis capabilities for Auto- Regressive 'AR' Integrated 'I' Moving-Average 'MA' models. In an AR1MA(p, q) model, p and q refer to the order of the Auto-Regressive and Moving-Average components of the model; d=1 refers to the order of differencing. Works by taking the first difference and delegating to the ARMA class. Also works for d=0 (no differencing).

The AR1MA class provides basic time-series analysis capabilities for Auto- Regressive 'AR' Integrated 'I' Moving-Average 'MA' models. In an AR1MA(p, q) model, p and q refer to the order of the Auto-Regressive and Moving-Average components of the model; d=1 refers to the order of differencing. Works by taking the first difference and delegating to the ARMA class. Also works for d=0 (no differencing).

Value parameters

diffr

whether to take a first difference (defaults to true)

hparam

the hyper-parameters

tt

the time vector, if relevant (time index may suffice)

y

the response vector (time-series data)

Attributes

Supertypes
trait Fit
trait FitM
trait Correlogram
trait Forecaster
trait Model
class Object
trait Matchable
class Any
Show all
object ARIMA

Companion object for class ARIMA. Includes features related to differencing and automated order selection.

Companion object for class ARIMA. Includes features related to differencing and automated order selection.

Attributes

See also
Companion
class
Supertypes
class Object
trait Matchable
class Any
Self type
ARIMA.type
class ARIMA(y: VectorD, tt: VectorD, hparam: HyperParameter) extends ARMA

The ARIMA class provides basic time series analysis capabilities for Auto- Regressive 'AR' Integrated 'I' Moving-Average 'MA' models. In an ARIMA(p, d, q) model, p and q refer to the order of the Auto-Regressive and Moving-Average components of the model; d refers to the order of differencing. Given time series data stored in vector y, its next value y_t = y(t) may be predicted based on prior values of y and its noise: y_t = δ + Σ(φ_i y_t-i) + Σ(θ_i e_t-i) + e_t where δ is a constant, φ is the auto-regressive coefficient vector, θ is the moving-average coefficient vector, and e is the noise vector.

If d > 0, then the time series must be differenced first before applying the above model.

Value parameters

hparam

the hyper-parameters

tt

the time vector, if relevant (time index may suffice)

y

the original input vector (time series data)

Attributes

Companion
object
Supertypes
class ARMA
trait Correlogram
trait Forecaster
trait Model
class FitI
trait Fit
trait FitM
class Object
trait Matchable
class Any
Show all
Known subtypes
class SARIMA
class SARIMAX
class ARMA(y: VectorD, tt: VectorD, hparam: HyperParameter) extends FitI, Forecaster, Correlogram

The ARMA class provides basic time series analysis capabilities for Auto-Regressive, Moving-Average (ARMA) models. In an ARMA(p, q) model, p refers to the order of the Auto-Regressive components and q refers to the Moving-Average compoenest of the model. ARMA models are often used for forecasting. Given time-series data stored in vector y, its next value y_t+1 = y(t+1) may be predicted based on prior values of y and its noise: y_t+1 = δ + Σ(φ_j y_t-j) + Σ(θ_j e_t-j) + e_t+1 where δ is a constant, φ is the auto-regressive coefficient vector, θ is the moving-average vector, and e_t+1 is the noise term.

The ARMA class provides basic time series analysis capabilities for Auto-Regressive, Moving-Average (ARMA) models. In an ARMA(p, q) model, p refers to the order of the Auto-Regressive components and q refers to the Moving-Average compoenest of the model. ARMA models are often used for forecasting. Given time-series data stored in vector y, its next value y_t+1 = y(t+1) may be predicted based on prior values of y and its noise: y_t+1 = δ + Σ(φ_j y_t-j) + Σ(θ_j e_t-j) + e_t+1 where δ is a constant, φ is the auto-regressive coefficient vector, θ is the moving-average vector, and e_t+1 is the noise term.

Value parameters

hparam

the hyper-parameters

tt

the time vector, if relevant (time index may suffice)

y

the response vector (time-series data)

Attributes

Companion
object
Supertypes
trait Correlogram
trait Forecaster
trait Model
class FitI
trait Fit
trait FitM
class Object
trait Matchable
class Any
Show all
Known subtypes
class ARIMA
class SARIMA
class SARIMAX
object ARMA

The ARMA companion object provides factory methods for the ARMA class.

The ARMA companion object provides factory methods for the ARMA class.

Attributes

Companion
class
Supertypes
class Object
trait Matchable
class Any
Self type
ARMA.type
class ARX(x: MatrixD, yy: VectorD, lags: Int, fname: Array[String], hparam: HyperParameter) extends Regression, ForecasterX

The ARX class supports regression for Time Series data. Multi-horizon forecasting supported via the RECURSIVE method. Given a response vector y, a predictor matrix x is built that consists of lagged y vectors,

The ARX class supports regression for Time Series data. Multi-horizon forecasting supported via the RECURSIVE method. Given a response vector y, a predictor matrix x is built that consists of lagged y vectors,

y_t = b dot x
where x = [1, y_{t-1}, y_{t-2}, ... y_{t-lags}]

Value parameters

fname

the feature/variable names

hparam

the hyper-parameters (use Regression.hp for default)

lags

the maximum lag included (inclusive)

x

the input/predictor matrix built out of lags of y (and optionally from exogenous variables ex)

yy

the output/response vector trimmed to match x.dim (@see ARX object)

Attributes

Companion
object
Supertypes
trait ForecasterX
class Regression
trait Fit
trait FitM
trait Predictor
trait Model
class Object
trait Matchable
class Any
Show all
object ARX

The ARX companion object provides factory methods.

The ARX companion object provides factory methods.

Attributes

Companion
class
Supertypes
class Object
trait Matchable
class Any
Self type
ARX.type
object ARX_MV

The ARX_MV object supports regression for Time Series data. Multi-horizon forecasting supported via the DIRECT method. Given a response vector y, a predictor matrix x is built that consists of lagged y vectors. Additional future response vectors are built for training. y_t = b dot x where x = [y_{t-1}, y_{t-2}, ... y_{t-lags}].

The ARX_MV object supports regression for Time Series data. Multi-horizon forecasting supported via the DIRECT method. Given a response vector y, a predictor matrix x is built that consists of lagged y vectors. Additional future response vectors are built for training. y_t = b dot x where x = [y_{t-1}, y_{t-2}, ... y_{t-lags}].

Attributes

Supertypes
class Object
trait Matchable
class Any
Self type
ARX_MV.type
class ARX_Quad(x: MatrixD, yy: VectorD, lags: Int, fname: Array[String], hparam: HyperParameter) extends Regression, ForecasterX

The ARX_Quad class supports quadratic regression for Time Series data. Given a response vector y, a predictor matrix x is built that consists of lagged y vectors,

The ARX_Quad class supports quadratic regression for Time Series data. Given a response vector y, a predictor matrix x is built that consists of lagged y vectors,

y_t = b dot x
where x = [ 1, y_{t-1}, y_{t-2}, ... y_{t-lag}, y_{t-1}^2, ...].

Value parameters

fname

the feature/variable names

hparam

the hyper-parameters (use Regression.hp for default)

lags

the maximum lag included (inclusive)

x

the input/predictor matrix built out of lags of y

yy

the output/response vector trimmed to match x.dim

Attributes

Companion
object
Supertypes
trait ForecasterX
class Regression
trait Fit
trait FitM
trait Predictor
trait Model
class Object
trait Matchable
class Any
Show all
object ARX_Quad

The ARX_Quad companion object provides factory methods.

The ARX_Quad companion object provides factory methods.

Attributes

Companion
class
Supertypes
class Object
trait Matchable
class Any
Self type
ARX_Quad.type
final class ARX_QuadTest

Attributes

Supertypes
class Object
trait Matchable
class Any
final class ARX_QuadTest2

Attributes

Supertypes
class Object
trait Matchable
class Any
final class ARX_QuadTest3

Attributes

Supertypes
class Object
trait Matchable
class Any
final class ARX_QuadTest4

Attributes

Supertypes
class Object
trait Matchable
class Any
final class ARX_QuadTest5

Attributes

Supertypes
class Object
trait Matchable
class Any
final class ARX_QuadTest6

Attributes

Supertypes
class Object
trait Matchable
class Any
object ARX_Quad_MV

The ARX_Quad_MV object supports quadratic regression for Time Series data. Multi-horizon forecasting supported via the DIRECT method. Given a response vector y, a predictor matrix x is built that consists of lagged y vectors. Additional future response vectors are built for training. y_t = b dot x where x = [y_{t-1}, y_{t-2}, ... y_{t-lags}]. Matrix x includes constant, linear and quadratic terms.

The ARX_Quad_MV object supports quadratic regression for Time Series data. Multi-horizon forecasting supported via the DIRECT method. Given a response vector y, a predictor matrix x is built that consists of lagged y vectors. Additional future response vectors are built for training. y_t = b dot x where x = [y_{t-1}, y_{t-2}, ... y_{t-lags}]. Matrix x includes constant, linear and quadratic terms.

Attributes

Supertypes
class Object
trait Matchable
class Any
Self type
object Example_Covid

The Example_Covid object provides a convenient way to load Covid-19 weekly data.

The Example_Covid object provides a convenient way to load Covid-19 weekly data.

Attributes

Supertypes
class Object
trait Matchable
class Any
Self type
object Example_GasFurnace

The Example_GasFurnace object provides a convenient way to load gas_furnace data.

The Example_GasFurnace object provides a convenient way to load gas_furnace data.

Attributes

Supertypes
class Object
trait Matchable
class Any
Self type
object Example_LakeLevels

The Example_LakeLevels provides a simple example time-series for testing forecasting models.

The Example_LakeLevels provides a simple example time-series for testing forecasting models.

Attributes

Supertypes
class Object
trait Matchable
class Any
Self type
trait Forecaster(y: VectorD, tt: VectorD, hparam: HyperParameter) extends Model

The Forecaster trait provides a common framework for several forecasters. Note, the train method must be called first followed by test.

The Forecaster trait provides a common framework for several forecasters. Note, the train method must be called first followed by test.

Value parameters

hparam

the hyper-parameters for models extending this trait

tt

the time vector, if relevant (index as time may suffice)

y

the response vector (time-series data)

Attributes

Companion
object
Supertypes
trait Model
class Object
trait Matchable
class Any
Known subtypes
class AR
class AR1MA
class ARMA
class ARIMA
class SARIMA
class SARIMAX
class NullModel
class QuadSpline
class RandomWalk
class TrendModel
Show all
object Forecaster

The Forecaster companion object provides functions useful for forecasting

The Forecaster companion object provides functions useful for forecasting

Attributes

Companion
trait
Supertypes
class Object
trait Matchable
class Any
Self type
Forecaster.type
trait ForecasterX(lags: Int)

The ForecasterX trait provides a common framework for several forecasting models that use 1 ENDOGENOUS variable y and 0 or more EXOGENOUS variables xj. It provides methods for multi-horizon (1 to h) forecasting using the RECURSIVE technique. Forecasted values are produced only for the endogenous variable y. Lower case indicates actual values, while upper case is for forecasted values.

The ForecasterX trait provides a common framework for several forecasting models that use 1 ENDOGENOUS variable y and 0 or more EXOGENOUS variables xj. It provides methods for multi-horizon (1 to h) forecasting using the RECURSIVE technique. Forecasted values are produced only for the endogenous variable y. Lower case indicates actual values, while upper case is for forecasted values.

Y_t+1 = f(y_t, y_t-1, ... y_t-p+1, x0_t, x0_t-1, ... x0_t-p+1, x1_t, ...) Y_t+2 = f(Y_t+1, y_t, ... y_t-p+2, x0_t, x0_t-1, ... x0_t-p+1, x1_t, ...) ... Y_t+h = f(Y_t+1, y_t, ... y_t-p+2, x0_t, x0_t-1, ... x0_t-p+1, x1_t, ...)

Value parameters

lags

the lags (p) used for endogenous variable (e.g., 10 => use lags 1 to 10)

Attributes

See also

Forecaster - when there are no exogenous variables

Supertypes
class Object
trait Matchable
class Any
Known subtypes
class ARX
class ARX_Quad
class KalmanFilter(x0: VectorD, ff: MatrixD, hh: MatrixD, qq: MatrixD, rr: MatrixD, gg: MatrixD, u: VectorD)

The KalmanFilter class is used to fit state-space models. x_t = F x_t-1 + G u_t + w_t (State Equation) z_t = H x_t + v_t (Observation/Measurement Equation)

The KalmanFilter class is used to fit state-space models. x_t = F x_t-1 + G u_t + w_t (State Equation) z_t = H x_t + v_t (Observation/Measurement Equation)

Value parameters

ff

the state transition matrix (F)

gg

the optional control-input matrix (G)

hh

the observation matrix (H)

qq

the process noise covariance matrix (Q)

rr

the observation noise covariance matrix (R)

u

the optional control vector

x0

the initial state vector

Attributes

Supertypes
class Object
trait Matchable
class Any
class NullModel(y: VectorD, tt: VectorD, hparam: HyperParameter) extends FitI, Forecaster, Correlogram

The NullModel class provides basic time series analysis capabilities for NullModel models. NullModel models are often used for forecasting. Given time series data stored in vector y, its next value y_t+1 = y(t+1) may be predicted based on prior value of y and its noise: y_t+1 = mu_y + e_t+1 where mu_y is the mean of y and e_t+1 is the noise term.

The NullModel class provides basic time series analysis capabilities for NullModel models. NullModel models are often used for forecasting. Given time series data stored in vector y, its next value y_t+1 = y(t+1) may be predicted based on prior value of y and its noise: y_t+1 = mu_y + e_t+1 where mu_y is the mean of y and e_t+1 is the noise term.

Value parameters

hparam

the hyper-parameters (none => use null)

tt

the time vector, if relevant (time index may suffice)

y

the response vector (time-series data)

Attributes

Companion
object
Supertypes
trait Correlogram
trait Forecaster
trait Model
class FitI
trait Fit
trait FitM
class Object
trait Matchable
class Any
Show all
object NullModel

The NullModel companion object provides factory methods for the NullModel class.

The NullModel companion object provides factory methods for the NullModel class.

Attributes

Companion
class
Supertypes
class Object
trait Matchable
class Any
Self type
NullModel.type
class QuadSpline(y: VectorD, tt: VectorD, hparam: HyperParameter) extends FitI, Forecaster, Correlogram

The QuadSpline class fits quadratic splines to time-series data that are equally spaced in time. A sliding window consisting of three data points is perfectly fit to a quadratic curve. y_t = a + bt + ct^2 Note, slope matching and smoothness issues are ignored.

The QuadSpline class fits quadratic splines to time-series data that are equally spaced in time. A sliding window consisting of three data points is perfectly fit to a quadratic curve. y_t = a + bt + ct^2 Note, slope matching and smoothness issues are ignored.

Value parameters

hparam

the hyper-parameters (none => use null)

tt

the time vector, if relevant (time index may suffice)

y

the response vector (time-series data)

Attributes

See also

wordsandbuttons.online/quadratic_splines_are_useful_too.html Any time point from t = 3 to the end of time series may be forecasted.

Companion
object
Supertypes
trait Correlogram
trait Forecaster
trait Model
class FitI
trait Fit
trait FitM
class Object
trait Matchable
class Any
Show all
object QuadSpline

The QuadSpline companion object provides factory methods for the QuadSpline class.

The QuadSpline companion object provides factory methods for the QuadSpline class.

Attributes

Companion
class
Supertypes
class Object
trait Matchable
class Any
Self type
QuadSpline.type
class RandomWalk(y: VectorD, tt: VectorD, hparam: HyperParameter) extends FitI, Forecaster, Correlogram

The RandomWalk class provides basic time series analysis capabilities for RandomWalk models. RandomWalk models are often used for forecasting. Given time series data stored in vector y, its next value y_t+1 = y(t+1) may be predicted based on prior value of y and its noise: y_t+1 = y_t + e_t+1 where e_t+1 is the noise vector.

The RandomWalk class provides basic time series analysis capabilities for RandomWalk models. RandomWalk models are often used for forecasting. Given time series data stored in vector y, its next value y_t+1 = y(t+1) may be predicted based on prior value of y and its noise: y_t+1 = y_t + e_t+1 where e_t+1 is the noise vector.

Value parameters

hparam

the hyper-parameters (none => use null)

tt

the time vector, if relevant (time index may suffice)

y

the response vector (time-series data)

Attributes

Companion
object
Supertypes
trait Correlogram
trait Forecaster
trait Model
class FitI
trait Fit
trait FitM
class Object
trait Matchable
class Any
Show all
object RandomWalk

The RandomWalk companion object provides factory methods for the RandomWalk class.

The RandomWalk companion object provides factory methods for the RandomWalk class.

Attributes

Companion
class
Supertypes
class Object
trait Matchable
class Any
Self type
RandomWalk.type
object RollingValidation

The RollingValidation object provides rolling-validation, where a full dataset is divided into a training set followed by a testing set. Retraining is done as the algorithm rolls through the testing set making out-of-sample predictions/forecasts to keep the parameters from becoming stale. For example, with TR_RATIO = 0.5 and m = 1000 it works as follows: tr(ain) 0 to 499, te(st) 500 to 999 Re-training occurs according to the retraining cycle rc, e.g., rc = 10 implies that retraining would occurs after every 10 forecasts or 50 times for this example.

The RollingValidation object provides rolling-validation, where a full dataset is divided into a training set followed by a testing set. Retraining is done as the algorithm rolls through the testing set making out-of-sample predictions/forecasts to keep the parameters from becoming stale. For example, with TR_RATIO = 0.5 and m = 1000 it works as follows: tr(ain) 0 to 499, te(st) 500 to 999 Re-training occurs according to the retraining cycle rc, e.g., rc = 10 implies that retraining would occurs after every 10 forecasts or 50 times for this example.

Attributes

Supertypes
class Object
trait Matchable
class Any
Self type
object SARIMA

Companion object for class SARIMA. Includes features related to differencing and automated order selection.

Companion object for class SARIMA. Includes features related to differencing and automated order selection.

Attributes

See also
Companion
class
Supertypes
class Object
trait Matchable
class Any
Self type
SARIMA.type
class SARIMA(y: VectorD, dd: Int, period: Int, tt: VectorD, hparam: HyperParameter) extends ARIMA

The SARIMA class provides basic time series analysis capabilities for Auto- Regressive AR Integrated I Moving-Average MA models. In a SARIMA(p, d, q) model, p and q refer to the order of the Auto-Regressive and Moving-Average components of the model; d refers to the order of differencing. Given time series data stored in vector y, its next value y_t+1 = y(t+1) may be predicted based on prior values of y and its noise: y_t+1 = δ + Σ(φ_i y_t-i) + Σ(θ_i e_t-i) + e_t+1 where δ is a constant, φ is the auto-regressive coefficient vector, θ is the moving-average coefficient vector, and e_t+1 is the noise vector.

If d > 0, then the time series must be differenced first before applying the above model. Seasonal differencing, auto-regressive and moving average factors can be incorporated into the model by applying seasonal differencing (possibly in addition to simple differencing) first, then add the seasonal auto-regressive and moving average terms, that rely on lagged values and errors, respectively, from one or more seasonal periods in the past, on the right hand side of the equation.

Value parameters

dd

the order of seasonal differencing

hparam

the hyper-parameters

period

the seasonal period (at least 2)

tt

the time vector, if relevant (time index may suffice)

y

the original input vector (time series data)

Attributes

Companion
object
Supertypes
class ARIMA
class ARMA
trait Correlogram
trait Forecaster
trait Model
class FitI
trait Fit
trait FitM
class Object
trait Matchable
class Any
Show all
Known subtypes
class SARIMAX
object SARIMAX

The SARIMAX companion object provides hyper-parameters for SARIMAX models and their subtypes.

The SARIMAX companion object provides hyper-parameters for SARIMAX models and their subtypes.

Attributes

Companion
class
Supertypes
class Object
trait Matchable
class Any
Self type
SARIMAX.type
class SARIMAX(y: VectorD, x: MatrixD, dd: Int, period: Int, tt: VectorD, hparam: HyperParameter) extends SARIMA

The SARIMAX class provides basic time series analysis capabilities for Auto- Regressive AR Integrated I Moving-Average MA, with eXogenous variables X models. In a SARIMAX(p, d, q)x(P, D, Q)_s [a, b] model, p and q are the orders of the AR and MA components; d is the order of differencing. P and Q are the orders of the AR and MA seasonal components; d is the order of Seasonal differencing. s is the seasonal period, and [a, b] is the range of lags for the eXogenous variables.

Value parameters

dd

the order of seasonal differencing

hparam

the hyper-parameters

period

the seasonal period (at least 2)

tt

the time vector, if relevant (time index may suffice)

x

the exogenous time series data as an input matrix

y

the original endogenous input vector (time series data)

Attributes

Companion
object
Supertypes
class SARIMA
class ARIMA
class ARMA
trait Correlogram
trait Forecaster
trait Model
class FitI
trait Fit
trait FitM
class Object
trait Matchable
class Any
Show all
class SimpleExpSmoothing(y: VectorD, tt: VectorD, hparam: HyperParameter) extends FitI, Forecaster, Correlogram

The SimpleExpSmoothing class provide very basic time series analysis using Simple Exponential Smoothing models. The forecasted value is the weighted average the latest value y_t and the latest smoothed value s_t. The smoothing parameter α in [0, 1] causes the contributions of older values to decay exponentially.

The SimpleExpSmoothing class provide very basic time series analysis using Simple Exponential Smoothing models. The forecasted value is the weighted average the latest value y_t and the latest smoothed value s_t. The smoothing parameter α in [0, 1] causes the contributions of older values to decay exponentially.

Value parameters

hparam

the hyper-parameters

tt

the time vector, if relevant (time index may suffice)

y

the response vector (original time series data)

Attributes

See also

Smoothing Equation in section 7.1. s_t+1 = α y_t + (1 - α) s_t smoothing equation yf_t+1 = s_t+1 forecast equation where vector s is the smoothed version of vector y.

Companion
object
Supertypes
trait Correlogram
trait Forecaster
trait Model
class FitI
trait Fit
trait FitM
class Object
trait Matchable
class Any
Show all
object SimpleExpSmoothing

The SimpleExpSmoothing companion object provides factory methods for the SimpleExpSmoothing class.

The SimpleExpSmoothing companion object provides factory methods for the SimpleExpSmoothing class.

Attributes

Companion
class
Supertypes
class Object
trait Matchable
class Any
Self type
class SimpleMovingAverage(y: VectorD, tt: VectorD, hparam: HyperParameter) extends FitI, Forecaster, Correlogram

The SimpleMovingAverage class provides basic time series analysis capabilities. For a SimpleMovingAverage model with the time series data stored in vector y, the next value y_t = y(t) may be predicted based on the mean of prior values of y and its noise: y_t+1 = mean (y_t, ..., y_t-q') + e_t+1 where e_t+1 is the noise vector and q' = q-1 the number of prior values used to compute the mean.

The SimpleMovingAverage class provides basic time series analysis capabilities. For a SimpleMovingAverage model with the time series data stored in vector y, the next value y_t = y(t) may be predicted based on the mean of prior values of y and its noise: y_t+1 = mean (y_t, ..., y_t-q') + e_t+1 where e_t+1 is the noise vector and q' = q-1 the number of prior values used to compute the mean.

Value parameters

hparam

the hyper-parameters

tt

the time points, if needed

y

the response vector (time series data)

Attributes

Companion
object
Supertypes
trait Correlogram
trait Forecaster
trait Model
class FitI
trait Fit
trait FitM
class Object
trait Matchable
class Any
Show all
object SimpleMovingAverage

The SimpleMovingAverage companion object provides factory methods for the SimpleMovingAverage class.

The SimpleMovingAverage companion object provides factory methods for the SimpleMovingAverage class.

Attributes

Companion
class
Supertypes
class Object
trait Matchable
class Any
Self type
object Stationarity_KPSS

The companion object for KPSS class, containing critical value coefficients needed in KPSS tests for Time Series Stationarity around a deterministic trend.

The companion object for KPSS class, containing critical value coefficients needed in KPSS tests for Time Series Stationarity around a deterministic trend.

Attributes

Companion
class
Supertypes
class Object
trait Matchable
class Any
Self type
class Stationarity_KPSS(yy: VectorD, lags_: Int, lagsType_: String, trend_: String) extends UnitRoot

The KPSS class provides capabilities of performing KPSS test to determine if a time series is stationary around a deterministic trend. This code is translated from the C++ code found in

The KPSS class provides capabilities of performing KPSS test to determine if a time series is stationary around a deterministic trend. This code is translated from the C++ code found in

Value parameters

lagsType_

type of lags, long or short

lags_

the number of lags to use

trend_

type of trend to test for

yy

the original time series vector

Attributes

See also

github.com/olmallet81/URT.

Companion
object
Supertypes
trait UnitRoot
class Object
trait Matchable
class Any
class TrendModel(y: VectorD, tt: VectorD, hparam: HyperParameter) extends FitI, Forecaster, Correlogram

The TrendModel class provides basic time series analysis capabilities for TrendModel models. TrendModel models are often used for forecasting. Given time series data stored in vector y, its next value y_t+1 = y(t+1) may be predicted based on time t and its noise: y_t+1 = b0 + b1 (t+1) + e_t+1 where b0 is the intercept, b1 is the slope and e_t+1 is the noise term.

The TrendModel class provides basic time series analysis capabilities for TrendModel models. TrendModel models are often used for forecasting. Given time series data stored in vector y, its next value y_t+1 = y(t+1) may be predicted based on time t and its noise: y_t+1 = b0 + b1 (t+1) + e_t+1 where b0 is the intercept, b1 is the slope and e_t+1 is the noise term.

Value parameters

hparam

the hyper-parameters (none => use null)

tt

the time vector (time index may suffice)

y

the response vector (time-series data)

Attributes

Companion
object
Supertypes
trait Correlogram
trait Forecaster
trait Model
class FitI
trait Fit
trait FitM
class Object
trait Matchable
class Any
Show all
object TrendModel

The TrendModel companion object provides factory methods for the TrendModel class.

The TrendModel companion object provides factory methods for the TrendModel class.

Attributes

Companion
class
Supertypes
class Object
trait Matchable
class Any
Self type
TrendModel.type
trait UnitRoot(val testName: String, val nobs: Int, val validTrends: VectorS, var lagsType: String, var lags: Int, var trend: String)

The UnitRoot trait provides a common framework for various unit root testers for Time Series Stationarity. This code is translated from the C++ code found in

The UnitRoot trait provides a common framework for various unit root testers for Time Series Stationarity. This code is translated from the C++ code found in

Value parameters

lags

the number of lags to use

lagsType

default lags value long or short time-series

nobs

the number of observations (length of time-series)

testName

the name of test, e.g., KPSS

trend

type of trend to test for

validTrends

vector of test valid trends types, e.g., constant, linear trend

Attributes

See also

github.com/olmallet81/URT.

Supertypes
class Object
trait Matchable
class Any
Known subtypes
object VAR

The VAR object supports regression for Multivariate Time Series data. Given a response matrix y, a predictor matrix x is built that consists of lagged y vectors. Additional future response vectors are built for training. y_t = b dot x where x = [y_{t-1}, y_{t-2}, ... y_{t-lag}].

The VAR object supports regression for Multivariate Time Series data. Given a response matrix y, a predictor matrix x is built that consists of lagged y vectors. Additional future response vectors are built for training. y_t = b dot x where x = [y_{t-1}, y_{t-2}, ... y_{t-lag}].

Attributes

Supertypes
class Object
trait Matchable
class Any
Self type
VAR.type
final class aR1MATest

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aR1MATest2

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aR1MATest3

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aR1MATest4

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRIMATest

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRIMATest2

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRIMATest3

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRIMATest4

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRMATest

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRMATest2

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRMATest3

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRMATest4

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRMATest5

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRMATest6

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRMATest7

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRTest

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRTest2

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRTest3

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRTest4

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRTest5

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRTest6

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRXTest

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRXTest2

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRXTest3

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRXTest4

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRXTest5

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRXTest6

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRXTest7

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRX_MVTest

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRX_MVTest2

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRX_MVTest3

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRX_MVTest4

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRX_MVTest5

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRX_MVTest6

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRX_Quad_MVTest

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRX_Quad_MVTest2

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRX_Quad_MVTest3

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRX_Quad_MVTest4

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRX_Quad_MVTest5

Attributes

Supertypes
class Object
trait Matchable
class Any
final class aRX_Quad_MVTest6

Attributes

Supertypes
class Object
trait Matchable
class Any
final class example_CovidTest

Attributes

Supertypes
class Object
trait Matchable
class Any
final class example_CovidTest2

Attributes

Supertypes
class Object
trait Matchable
class Any
final class example_GasFurnaceTest

Attributes

Supertypes
class Object
trait Matchable
class Any
final class example_GasFurnaceTest2

Attributes

Supertypes
class Object
trait Matchable
class Any
final class kalmanFilterTest

Attributes

Supertypes
class Object
trait Matchable
class Any
final class nullModelTest

Attributes

Supertypes
class Object
trait Matchable
class Any
final class nullModelTest2

Attributes

Supertypes
class Object
trait Matchable
class Any
final class nullModelTest3

Attributes

Supertypes
class Object
trait Matchable
class Any
final class quadSplineTest

Attributes

Supertypes
class Object
trait Matchable
class Any
final class quadSplineTest2

Attributes

Supertypes
class Object
trait Matchable
class Any
final class quadSplineTest3

Attributes

Supertypes
class Object
trait Matchable
class Any
final class quadSplineTest4

Attributes

Supertypes
class Object
trait Matchable
class Any
final class randomWalkTest

Attributes

Supertypes
class Object
trait Matchable
class Any
final class randomWalkTest2

Attributes

Supertypes
class Object
trait Matchable
class Any
final class randomWalkTest3

Attributes

Supertypes
class Object
trait Matchable
class Any
final class randomWalkTest4

Attributes

Supertypes
class Object
trait Matchable
class Any
final class rollingValidationTest

Attributes

Supertypes
class Object
trait Matchable
class Any
final class rollingValidationTest2

Attributes

Supertypes
class Object
trait Matchable
class Any
final class rollingValidationTest3

Attributes

Supertypes
class Object
trait Matchable
class Any
final class rollingValidationTest4

Attributes

Supertypes
class Object
trait Matchable
class Any
final class sARIMATest

Attributes

Supertypes
class Object
trait Matchable
class Any
final class sARIMATest2

Attributes

Supertypes
class Object
trait Matchable
class Any
final class sARIMATest3

Attributes

Supertypes
class Object
trait Matchable
class Any
final class sARIMATest4

Attributes

Supertypes
class Object
trait Matchable
class Any
final class sARIMATest5

Attributes

Supertypes
class Object
trait Matchable
class Any
final class sARIMATest6

Attributes

Supertypes
class Object
trait Matchable
class Any
final class sARIMAXTest

Attributes

Supertypes
class Object
trait Matchable
class Any
final class sARIMAXTest2

Attributes

Supertypes
class Object
trait Matchable
class Any
final class sARIMAXTest3

Attributes

Supertypes
class Object
trait Matchable
class Any
final class sARIMAXTest4

Attributes

Supertypes
class Object
trait Matchable
class Any
final class sARIMAXTest5

Attributes

Supertypes
class Object
trait Matchable
class Any
final class sARIMAXTest6

Attributes

Supertypes
class Object
trait Matchable
class Any
final class simpleExpSmoothingTest

Attributes

Supertypes
class Object
trait Matchable
class Any
final class simpleExpSmoothingTest2

Attributes

Supertypes
class Object
trait Matchable
class Any
final class simpleExpSmoothingTest3

Attributes

Supertypes
class Object
trait Matchable
class Any
final class simpleExpSmoothingTest4

Attributes

Supertypes
class Object
trait Matchable
class Any
final class simpleExpSmoothingTest5

Attributes

Supertypes
class Object
trait Matchable
class Any
final class simpleMovingAverageTest

Attributes

Supertypes
class Object
trait Matchable
class Any
final class simpleMovingAverageTest2

Attributes

Supertypes
class Object
trait Matchable
class Any
final class simpleMovingAverageTest3

Attributes

Supertypes
class Object
trait Matchable
class Any
final class simpleMovingAverageTest4

Attributes

Supertypes
class Object
trait Matchable
class Any
final class simpleMovingAverageTest5

Attributes

Supertypes
class Object
trait Matchable
class Any
final class stationarity_KPSSTest

Attributes

Supertypes
class Object
trait Matchable
class Any
final class stationarity_KPSSTest2

Attributes

Supertypes
class Object
trait Matchable
class Any
final class stationarity_KPSSTest3

Attributes

Supertypes
class Object
trait Matchable
class Any
final class stationaryTest

Attributes

Supertypes
class Object
trait Matchable
class Any
final class stationaryTest2

Attributes

Supertypes
class Object
trait Matchable
class Any
final class stationaryTest3

Attributes

Supertypes
class Object
trait Matchable
class Any
final class trendModelTest

Attributes

Supertypes
class Object
trait Matchable
class Any
final class trendModelTest2

Attributes

Supertypes
class Object
trait Matchable
class Any
final class trendModelTest3

Attributes

Supertypes
class Object
trait Matchable
class Any
final class varTest

Attributes

Supertypes
class Object
trait Matchable
class Any
final class varTest2

Attributes

Supertypes
class Object
trait Matchable
class Any
final class varTest3

Attributes

Supertypes
class Object
trait Matchable
class Any
final class varTest4

Attributes

Supertypes
class Object
trait Matchable
class Any
final class varTest5

Attributes

Supertypes
class Object
trait Matchable
class Any
final class varTest6

Attributes

Supertypes
class Object
trait Matchable
class Any

Types

type CriticalValues = HashMap[Int, VectorD]

Value members

Concrete methods

def ARX_QuadTest(): Unit

The ARX_QuadTest main function tests the ARX_Quad class. This test is used to CHECK that the buildMatrix4TS function is working correctly. May get NaN for some maximum lags (p) due to multi-collinearity.

The ARX_QuadTest main function tests the ARX_Quad class. This test is used to CHECK that the buildMatrix4TS function is working correctly. May get NaN for some maximum lags (p) due to multi-collinearity.

runMain scalation.modeling.forecasting.ARX_QuadTest

Attributes

def ARX_QuadTest2(): Unit

The ARX_QuadTest2 main function tests the ARX_Quad class on real data: Forecasting lake levels.

The ARX_QuadTest2 main function tests the ARX_Quad class on real data: Forecasting lake levels.

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.ARX_QuadTest2

def ARX_QuadTest3(): Unit

The ARX_QuadTest3 main function tests the ARX_Quad class on real data: Forecasting COVID-19. Does In-Sample Testing on Endogenous variable.

The ARX_QuadTest3 main function tests the ARX_Quad class on real data: Forecasting COVID-19. Does In-Sample Testing on Endogenous variable.

runMain scalation.modeling.forecasting.ARX_QuadTest3

Attributes

def ARX_QuadTest4(): Unit

The ARX_QuadTest4 main function tests the ARX_Quad class on real data: Forecasting COVID-19 Weekly Data. Does In-Sample Testing on endogenous and exogenous variables. Stepsise gives: 0, 19, 40, 37, 60, 17, 15, 53, 5, 20, 50, 49, 48, 47, 18, 9, 6 best R^2-bar Stepsise gives: 0, 19, 40, 37, 60, 17, 15, 53, 5, 20, 50, 49, 48, 47 best sMAPE

The ARX_QuadTest4 main function tests the ARX_Quad class on real data: Forecasting COVID-19 Weekly Data. Does In-Sample Testing on endogenous and exogenous variables. Stepsise gives: 0, 19, 40, 37, 60, 17, 15, 53, 5, 20, 50, 49, 48, 47, 18, 9, 6 best R^2-bar Stepsise gives: 0, 19, 40, 37, 60, 17, 15, 53, 5, 20, 50, 49, 48, 47 best sMAPE

runMain scalation.modeling.forecasting.ARX_QuadTest4

Attributes

def ARX_QuadTest5(): Unit

The ARX_QuadTest5 main function tests the ARX_Quad class on real data: Forecasting COVID-19 Weekly Data. Does TnT Testing on endogenous and exogenous variables. Determine the terms to include in the model for TnT from using Stepwise on In-Sample. Stepsise gives: 0, 19, 40, 37, 60, 17, 15, 53, 5, 20, 50, 49, 48, 47, 18, 9, 6 best R^2-bar Stepsise gives: 0, 19, 40, 37, 60, 17, 15, 53, 5, 20, 50, 49, 48, 47 best sMAPE

The ARX_QuadTest5 main function tests the ARX_Quad class on real data: Forecasting COVID-19 Weekly Data. Does TnT Testing on endogenous and exogenous variables. Determine the terms to include in the model for TnT from using Stepwise on In-Sample. Stepsise gives: 0, 19, 40, 37, 60, 17, 15, 53, 5, 20, 50, 49, 48, 47, 18, 9, 6 best R^2-bar Stepsise gives: 0, 19, 40, 37, 60, 17, 15, 53, 5, 20, 50, 49, 48, 47 best sMAPE

runMain scalation.modeling.forecasting.ARX_QuadTest5

Attributes

def ARX_QuadTest6(): Unit

The ARX_QuadTest6 main function tests the ARX_Quad class on real data: Forecasting COVID-19 Weekly Data. Does Rolling Validation on endogenous and exogenous variables. Determine the terms to include in the model HOW?

The ARX_QuadTest6 main function tests the ARX_Quad class on real data: Forecasting COVID-19 Weekly Data. Does Rolling Validation on endogenous and exogenous variables. Determine the terms to include in the model HOW?

runMain scalation.modeling.forecasting.ARX_QuadTest6

Attributes

def aR1MATest(): Unit

The aR1MATest main function tests the AR1MA class on simulated data. Test predictions (one step ahead forecasts).

The aR1MATest main function tests the AR1MA class on simulated data. Test predictions (one step ahead forecasts).

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.aR1MATest

def aR1MATest2(): Unit

The aR1MATest2 main function tests the AR1MA class on real data: Forecasting lake levels. Test predictions (one step ahead forecasts) with no differencing

The aR1MATest2 main function tests the AR1MA class on real data: Forecasting lake levels. Test predictions (one step ahead forecasts) with no differencing

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.aR1MATest2

def aR1MATest3(): Unit

The aR1MATest3 main function tests the AR1MA class on real data: Forecasting lake levels. Test predictions (one step ahead forecasts) taking one difference.

The aR1MATest3 main function tests the AR1MA class on real data: Forecasting lake levels. Test predictions (one step ahead forecasts) taking one difference.

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.aR1MATest3

def aR1MATest4(): Unit

The aR1MATest4 main function tests the AR1MA class on real data: Forecasting COVID-19.

The aR1MATest4 main function tests the AR1MA class on real data: Forecasting COVID-19.

runMain scalation.modeling.forecasting.aR1MATest4

Attributes

def aRIMATest(): Unit

The aRIMATest main function tests the ARIMA class on real data: Forecasting lake levels.

The aRIMATest main function tests the ARIMA class on real data: Forecasting lake levels.

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.aRIMATest

def aRIMATest2(): Unit

The aRIMATest2 main function tests the ARIMA class. Test simulated data.

The aRIMATest2 main function tests the ARIMA class. Test simulated data.

runMain scalation.modeling.forecasting.aRIMATest2

Attributes

def aRIMATest3(): Unit

The aRIMATest3 main function tests the ARIMA class. Traffic dataset.

The aRIMATest3 main function tests the ARIMA class. Traffic dataset.

runMain scalation.modeling.forecasting.aRIMATest3

Attributes

def aRIMATest4(): Unit

The aRIMATest4 main function tests the ARIMA class. Simulated data with a quadratic pattern.

The aRIMATest4 main function tests the ARIMA class. Simulated data with a quadratic pattern.

runMain scalation.modeling.forecasting.aRIMATest4

Attributes

def aRMATest(): Unit

The aRMATest main function tests the ARMA class on simulated data. Test predictions (one step ahead forecasts).

The aRMATest main function tests the ARMA class on simulated data. Test predictions (one step ahead forecasts).

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.aRTest

def aRMATest2(): Unit

The aRMATest2 main function tests the ARMA class on real data: Forecasting lake levels. Test predictions (one step ahead forecasts).

The aRMATest2 main function tests the ARMA class on real data: Forecasting lake levels. Test predictions (one step ahead forecasts).

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.aRMATest2

def aRMATest3(): Unit

The aRMATest3 main function tests the ARMA class on real data: Forecasting lake levels. Test forecasts (1 to h steps ahead forecasts).

The aRMATest3 main function tests the ARMA class on real data: Forecasting lake levels. Test forecasts (1 to h steps ahead forecasts).

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.aRMATest3

def aRMATest4(): Unit

The aRMATest4 main function tests the ARMA class on real data: Forecasting lake levels.

The aRMATest4 main function tests the ARMA class on real data: Forecasting lake levels.

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.aRMATest4

def aRMATest5(): Unit

The aRMATest5 main function tests the ARMA class on real data: Forecasting lake levels. This test looks at the velocity series (first differences). Test predictions (one step ahead forecasts).

The aRMATest5 main function tests the ARMA class on real data: Forecasting lake levels. This test looks at the velocity series (first differences). Test predictions (one step ahead forecasts).

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.aRMATest5

def aRMATest6(): Unit

The aRMATest6 main function tests the ARMA class on real data: Forecasting COVID-19.

The aRMATest6 main function tests the ARMA class on real data: Forecasting COVID-19.

runMain scalation.modeling.forecasting.aRMATest6

Attributes

def aRMATest7(): Unit

The aRMATest6 main function tests the AR class on real data: Forecasting Weekly Covid-19. Test forecasts (1 to h steps ahead forecasts). Try multiple values for p.

The aRMATest6 main function tests the AR class on real data: Forecasting Weekly Covid-19. Test forecasts (1 to h steps ahead forecasts). Try multiple values for p.

runMain scalation.modeling.forecasting.aRMATest7

Attributes

def aRTest(): Unit

The aRTest main function tests the AR class on simulated data. Test predictions (one step ahead forecasts).

The aRTest main function tests the AR class on simulated data. Test predictions (one step ahead forecasts).

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.aRTest

def aRTest2(): Unit

The aRTest2 main function tests the AR class on real data: Forecasting lake levels. Test predictions (one step ahead forecasts).

The aRTest2 main function tests the AR class on real data: Forecasting lake levels. Test predictions (one step ahead forecasts).

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.aRTest2

def aRTest3(): Unit

The aRTest3 main function tests the AR class on real data: Forecasting lake levels. Test forecasts (1 to h steps ahead forecasts).

The aRTest3 main function tests the AR class on real data: Forecasting lake levels. Test forecasts (1 to h steps ahead forecasts).

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.aRTest3

def aRTest4(): Unit

The aRTest4 main function tests the AR class on real data: Forecasting lake levels. Test forecasts (1 to h steps ahead forecasts). Try multiple values for p.

The aRTest4 main function tests the AR class on real data: Forecasting lake levels. Test forecasts (1 to h steps ahead forecasts). Try multiple values for p.

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.aRTest4

def aRTest5(): Unit

The aRTest5 main function tests the AR class on real data: Forecasting Weekly Covid-19. Test forecasts (1 to h steps ahead forecasts). Try multiple values for p.

The aRTest5 main function tests the AR class on real data: Forecasting Weekly Covid-19. Test forecasts (1 to h steps ahead forecasts). Try multiple values for p.

runMain scalation.modeling.forecasting.aRTest5

Attributes

def aRTest6(): Unit

The aRTest6 main function tests the AR class on simulated data. Test predictions (one step ahead forecasts).

The aRTest6 main function tests the AR class on simulated data. Test predictions (one step ahead forecasts).

runMain scalation.modeling.forecasting.aRTest6

Attributes

def aRXTest(): Unit

The aRXTest main function tests the ARX class. This test is used to CHECK that the buildMatrix4TS function is working correctly. May get NaN for some maximum lags (p) due to multi-collinearity.

The aRXTest main function tests the ARX class. This test is used to CHECK that the buildMatrix4TS function is working correctly. May get NaN for some maximum lags (p) due to multi-collinearity.

runMain scalation.modeling.forecasting.aRXTest

Attributes

def aRXTest2(): Unit

The aRXTest2 main function tests the ARX class on real data: Forecasting lake levels.

The aRXTest2 main function tests the ARX class on real data: Forecasting lake levels.

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.aRXTest2

def aRXTest3(): Unit

The aRXTest3 main function tests the ARX class on real data: Forecasting COVID-19 Daily Data. Does In-Sample Testing on Endogenous variable.

The aRXTest3 main function tests the ARX class on real data: Forecasting COVID-19 Daily Data. Does In-Sample Testing on Endogenous variable.

runMain scalation.modeling.forecasting.aRXTest3

Attributes

def aRXTest4(): Unit

The aRXTest4 main function tests the ARX class on real data: Forecasting COVID-19 Weekly Data. Does In-Sample Testing on endogenous and exogenous variables.

The aRXTest4 main function tests the ARX class on real data: Forecasting COVID-19 Weekly Data. Does In-Sample Testing on endogenous and exogenous variables.

runMain scalation.modeling.forecasting.aRXTest4

Attributes

def aRXTest5(): Unit

The aRXTest5 main function tests the ARX class on real data: Forecasting COVID-19 Weekly Data. Does TnT Testing on endogenous and exogenous variables. Determine the terms to include in the model using Stepwise on In-Sample.

The aRXTest5 main function tests the ARX class on real data: Forecasting COVID-19 Weekly Data. Does TnT Testing on endogenous and exogenous variables. Determine the terms to include in the model using Stepwise on In-Sample.

runMain scalation.modeling.forecasting.aRXTest5

Attributes

def aRXTest6(): Unit

The aRXTest6 main function tests the ARX class on real data: Forecasting COVID-19 Weekly Data. Does Rolling Validation on endogenous and exogenous variables. Determine the terms to include in the model using Stepwise on In-Sample.

The aRXTest6 main function tests the ARX class on real data: Forecasting COVID-19 Weekly Data. Does Rolling Validation on endogenous and exogenous variables. Determine the terms to include in the model using Stepwise on In-Sample.

runMain scalation.modeling.forecasting.aRXTest6

Attributes

def aRXTest7(): Unit

The aRXTest7 main function tests the ARX class on real data: Forecasting COVID-19 Weekly Data. Preliminary investigation of Symbolic Regression.

The aRXTest7 main function tests the ARX class on real data: Forecasting COVID-19 Weekly Data. Preliminary investigation of Symbolic Regression.

runMain scalation.modeling.forecasting.aRXTest7

Attributes

def aRX_MVTest(): Unit

The aRX_MVTest main function tests the ARX_MV class. This test is used to CHECK that the buildMatrix4TS function is working correctly. May get NaN for some maximum lags (p) due to multi-collinearity.

The aRX_MVTest main function tests the ARX_MV class. This test is used to CHECK that the buildMatrix4TS function is working correctly. May get NaN for some maximum lags (p) due to multi-collinearity.

runMain scalation.modeling.forecasting.aRX_MVTest

Attributes

def aRX_MVTest2(): Unit

The aRX_MVTest2 main function tests the ARX_MV class on real data: Forecasting lake levels.

The aRX_MVTest2 main function tests the ARX_MV class on real data: Forecasting lake levels.

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.aRX_MVTest2

def aRX_MVTest3(): Unit

The aRX_MVTest3 main function tests the ARX_MV class on real data: Forecasting COVID-19 Weekly Data. Performs In-Sample Testing using endogenous variable.

The aRX_MVTest3 main function tests the ARX_MV class on real data: Forecasting COVID-19 Weekly Data. Performs In-Sample Testing using endogenous variable.

runMain scalation.modeling.forecasting.aRX_MVTest3

Attributes

def aRX_MVTest4(): Unit

The aRX_MVTest4 main function tests the ARX_MV class on real data: Forecasting COVID-19 Weekly Data.

The aRX_MVTest4 main function tests the ARX_MV class on real data: Forecasting COVID-19 Weekly Data.

runMain scalation.modeling.forecasting.aRX_MVTest4

Attributes

def aRX_MVTest5(): Unit

The aRX_MVTest5 main function tests the ARX_MV class on real data: Forecasting COVID-19 Weekly Data. Does TnT Testing on endogenous and exogenous variables. Determine the terms to include in the model using Stepwise on In-Sample.

The aRX_MVTest5 main function tests the ARX_MV class on real data: Forecasting COVID-19 Weekly Data. Does TnT Testing on endogenous and exogenous variables. Determine the terms to include in the model using Stepwise on In-Sample.

runMain scalation.modeling.forecasting.aRX_MVTest5

Attributes

def aRX_MVTest6(): Unit

The aRX_MVTest6 main function tests the ARX_MV class on real data: Forecasting COVID-19 Weekly Data. Does Rolling Validation on endogenous and exogenous variables. Determine the terms to include in the model using Stepwise on In-Sample.

The aRX_MVTest6 main function tests the ARX_MV class on real data: Forecasting COVID-19 Weekly Data. Does Rolling Validation on endogenous and exogenous variables. Determine the terms to include in the model using Stepwise on In-Sample.

runMain scalation.modeling.forecasting.aRX_MVTest6

Attributes

def aRX_Quad_MVTest(): Unit

The aRX_Quad_MVTest main function tests the ARX_Quad_MV object. This test is used to CHECK that the buildMatrix4TS function is working correctly. May get NaN for some maximum lags (p) due to multi-collinearity.

The aRX_Quad_MVTest main function tests the ARX_Quad_MV object. This test is used to CHECK that the buildMatrix4TS function is working correctly. May get NaN for some maximum lags (p) due to multi-collinearity.

runMain scalation.modeling.forecasting.aRX_Quad_MVTest

Attributes

def aRX_Quad_MVTest2(): Unit

The aRX_Quad_MVTest2 main function tests the ARX_Quad_MV class on real data: Forecasting lake levels. Uses quadratic regression.

The aRX_Quad_MVTest2 main function tests the ARX_Quad_MV class on real data: Forecasting lake levels. Uses quadratic regression.

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.aRX_Quad_MVTest2

def aRX_Quad_MVTest3(): Unit

The aRX_Quad_MVTest3 main function tests the ARX_Quad_MV class on real data: Forecasting COVID-19 Weekly Data. Uses quadratic regression, In-Sample Testing using endogenous variable.

The aRX_Quad_MVTest3 main function tests the ARX_Quad_MV class on real data: Forecasting COVID-19 Weekly Data. Uses quadratic regression, In-Sample Testing using endogenous variable.

runMain scalation.modeling.forecasting.aRX_Quad_MVTest3

Attributes

def aRX_Quad_MVTest4(): Unit

The aRX_Quad_MVTest4 main function tests the ARX_Quad_MV class on real data: Forecasting COVID-19 Weekly Data. Uses quadratic regression, In-Sample Testing using endogenous and exogeneous variables.

The aRX_Quad_MVTest4 main function tests the ARX_Quad_MV class on real data: Forecasting COVID-19 Weekly Data. Uses quadratic regression, In-Sample Testing using endogenous and exogeneous variables.

runMain scalation.modeling.forecasting.aRX_Quad_MVTest4

Attributes

def aRX_Quad_MVTest5(): Unit

The aRX_Quad_MVTest5 main function tests the ARX_Quad_MV class on real data: Forecasting COVID-19 Weekly Data. Uses Quadratic Regression. Does TnT Testing on endogenous and exogenous variables. Determine the terms to include in the model for TnT from using Stepwise on In-Sample.

The aRX_Quad_MVTest5 main function tests the ARX_Quad_MV class on real data: Forecasting COVID-19 Weekly Data. Uses Quadratic Regression. Does TnT Testing on endogenous and exogenous variables. Determine the terms to include in the model for TnT from using Stepwise on In-Sample.

runMain scalation.modeling.forecasting.aRX_Quad_MVTest5

Attributes

def aRX_Quad_MVTest6(): Unit

The aRX_Quad_MVTest6 main function tests the ARX_Quad_MV class on real data: Forecasting COVID-19 Weekly Data. Uses Quadratic Regression. Does TnT Testing on endogenous and exogenous variables. Determine the terms to include in the model for TnT from using Stepwise on In-Sample.

The aRX_Quad_MVTest6 main function tests the ARX_Quad_MV class on real data: Forecasting COVID-19 Weekly Data. Uses Quadratic Regression. Does TnT Testing on endogenous and exogenous variables. Determine the terms to include in the model for TnT from using Stepwise on In-Sample.

runMain scalation.modeling.forecasting.aRX_Quad_MVTest6

Attributes

def buildMatrix4TS(y: VectorD, lags: Int, h: Int): (MatrixD, MatrixD)

Given a response vector y, build and return (1) an input/predictor MATRIX xx and (2) an output/multi-horizon output/response MATRIX yy. Used by Multi-Variate (MV) forecast models such as ARX_MV. that use DIRECT multi-horizon forecasting. The first lag responses can't be predicted due to missing past values. The last h-1 responses can't be predicted due to missing future values. Therefore the number of rows in xx and yy is reduced to y.dim + 1 - lags - h.

Given a response vector y, build and return (1) an input/predictor MATRIX xx and (2) an output/multi-horizon output/response MATRIX yy. Used by Multi-Variate (MV) forecast models such as ARX_MV. that use DIRECT multi-horizon forecasting. The first lag responses can't be predicted due to missing past values. The last h-1 responses can't be predicted due to missing future values. Therefore the number of rows in xx and yy is reduced to y.dim + 1 - lags - h.

FIX - try to extend to "val xx = new MatrixD (y.dim - h, lags)"

Value parameters

h

the forecasting horizon (1, 2, ... h)

lags

the maximum lag included (inclusive)

y

the given output/response vector

Attributes

def buildMatrix4TS(y: VectorD, lags: Int): (MatrixD, VectorD)

Given a response vector y, build and return (1) an input/predictor MATRIX xx and (2) an output/single-horizon output/response VECTOR yy. Used by Single-Variate forecast models such as ARX. that use RECURSIVE multi-horizon forecasting. The first response can't be predicted due to missing past values. Therefore the number of rows in xx and yy is reduced to y.dim - 1 (day 0 cut out).

Given a response vector y, build and return (1) an input/predictor MATRIX xx and (2) an output/single-horizon output/response VECTOR yy. Used by Single-Variate forecast models such as ARX. that use RECURSIVE multi-horizon forecasting. The first response can't be predicted due to missing past values. Therefore the number of rows in xx and yy is reduced to y.dim - 1 (day 0 cut out).

Value parameters

lags

the maximum lag included (inclusive)

y

the given output/response vector

Attributes

def buildMatrix4TS_exo(ex: VectorD, lags: Int, elag1: Int, elag2: Int): MatrixD

Given an exogenous variable vector ex corresponding to an endogenous response vector y, build and return an input/predictor MATRIX xx. The first lag responses can't be predicted due to missing past values. Therefore the number of rows in xx is reduced to ex.dim - elag1.

Given an exogenous variable vector ex corresponding to an endogenous response vector y, build and return an input/predictor MATRIX xx. The first lag responses can't be predicted due to missing past values. Therefore the number of rows in xx is reduced to ex.dim - elag1.

Value parameters

elag1

the minimum lag included (inclusive) for the exogenous variable

elag2

the maximum lag included (inclusive) for the exogenous variable

ex

the exogenous variable vector

lags

the maximum lag included (inclusive) for the endogenous variable

Attributes

def buildTensor4TS(y: VectorD, ex: MatrixD, lags: Int, h: Int)(el: Int): (TensorD, MatrixD)

Build a 3D tensor that collects and (lag) expands the endogenous variable and all exogenous variables into the form 'time x lags x variables' --> INPUT TENSOR. Lags for endogenous variable: 1, 2, ... lags Lags for exogenous variables: el, el+1, ... el-1+lags For models like SARIMAX with weekly data, exo variables are not forecasted, so it is not possible to use exogenous lag 1 to make week-two forecasts. Also build a matrix of target values for each forecasting horizon --> OUTPUT MATRIX. The number of rows m = y.dim - el, as forecasts cannot be made unless there is at least one endogenous and one exogenous lag (past value) available to the model. Model: yy_hat = f(xx) with loss function, e.g., || yy - yy_hat ||_F NOTE: for models not taking tensor input, flatten into a matrix.

Build a 3D tensor that collects and (lag) expands the endogenous variable and all exogenous variables into the form 'time x lags x variables' --> INPUT TENSOR. Lags for endogenous variable: 1, 2, ... lags Lags for exogenous variables: el, el+1, ... el-1+lags For models like SARIMAX with weekly data, exo variables are not forecasted, so it is not possible to use exogenous lag 1 to make week-two forecasts. Also build a matrix of target values for each forecasting horizon --> OUTPUT MATRIX. The number of rows m = y.dim - el, as forecasts cannot be made unless there is at least one endogenous and one exogenous lag (past value) available to the model. Model: yy_hat = f(xx) with loss function, e.g., || yy - yy_hat ||_F NOTE: for models not taking tensor input, flatten into a matrix.

Value parameters

el

the first exogenous lag (may be larger than 1)

ex

the exogenous variable matrix over time x exo_vars (e.g., icu_patients, hosp_patient)

h

the forecasting horizon (1, 2, ... h)

lags

the maximum lag included (inclusive)

y

the endogenous variable vector over time (e.g., new_deaths)

Attributes

def del(y: VectorD): VectorD

Return the first difference of the time-series y, giving the velocity v_t = y_t+1 - y_t.

Return the first difference of the time-series y, giving the velocity v_t = y_t+1 - y_t.

Value parameters

y

the original time-series to be differenced

Attributes

def example_CovidTest(): Unit

The example_CovidTest main function test the Example_Covid object.

The example_CovidTest main function test the Example_Covid object.

runMain scalation.modeling.forecasting.example_CovidTest

Attributes

def example_CovidTest2(): Unit

The example_CovidTest2 main function test the Example_Covid object. This performs Exploratory Data Analysis (EDA) to find relationships between contemporaneous variables.

The example_CovidTest2 main function test the Example_Covid object. This performs Exploratory Data Analysis (EDA) to find relationships between contemporaneous variables.

runMain scalation.modeling.forecasting.example_CovidTest2

Attributes

def example_GasFurnaceTest(): Unit

The example_GasFurnaceTest main function test the Example_GasFurnace object.

The example_GasFurnaceTest main function test the Example_GasFurnace object.

runMain scalation.modeling.forecasting.example_GasFurnaceTest

Attributes

def example_GasFurnaceTest2(): Unit

The example_GasFurnaceTest2 main function test the Example_GasFurnace object. This performs Exploratory Data Analysis (EDA) to find relationships between contemporaneous variables.

The example_GasFurnaceTest2 main function test the Example_GasFurnace object. This performs Exploratory Data Analysis (EDA) to find relationships between contemporaneous variables.

runMain scalation.modeling.forecasting.example_GasFurnaceTest2

Attributes

def kalmanFilterTest(): Unit

The kalmanFilterTest main function is used to test the KalmanFilter class.

The kalmanFilterTest main function is used to test the KalmanFilter class.

Attributes

See also

en.wikipedia.org/wiki/Kalman_filter

runMain scalation.modeling.forecasting.kalmanFilterTest

def makeTSeries(signal: FunctionS2S, m: Int, noise: Variate): VectorD

The makeTSeries top level function generates time-series data.

The makeTSeries top level function generates time-series data.

Value parameters

m

the length of the time series

noise

the random variate generator used for the noise part

signal

the function of time used to make the deterministic part

Attributes

def makeTSeriesR(c: Double, φ: VectorD, m: Int, noise: Variate): VectorD

The makeTSeries top level function recursively generates time-series data by simulating and AR process. y_t+1 = δ + Σ(φ_j y_t-j) + e_t+1 Note: all defaults generates white noise with variance 1

The makeTSeries top level function recursively generates time-series data by simulating and AR process. y_t+1 = δ + Σ(φ_j y_t-j) + e_t+1 Note: all defaults generates white noise with variance 1

Value parameters

c

the initial value for the time series

m

the length of the time series

noise

the random variate generator used for the noise part

φ

the auto-regressive coefficients

Attributes

def nullModelTest(): Unit

The nullModelTest main function tests the NullModel class on simulated data. Test predictions (one step ahead forecasts).

The nullModelTest main function tests the NullModel class on simulated data. Test predictions (one step ahead forecasts).

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.nullModelTest

def nullModelTest2(): Unit

The nullModelTest2 main function tests the NullModel class on real data: Forecasting lake levels. Test predictions (one step ahead forecasts).

The nullModelTest2 main function tests the NullModel class on real data: Forecasting lake levels. Test predictions (one step ahead forecasts).

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.nullModelTest2

def nullModelTest3(): Unit

The nullModelTest3 main function tests the NullModel class on real data: Forecasting lake levels. Test forecasts (1 to h steps ahead forecasts).

The nullModelTest3 main function tests the NullModel class on real data: Forecasting lake levels. Test forecasts (1 to h steps ahead forecasts).

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.nullModelTest3

def pq(hpar: HyperParameter): Int
def quadSplineTest(): Unit

The quadSplineTest main function tests the QuadSpline class on simulated data. Test predictions (one step ahead forecasts).

The quadSplineTest main function tests the QuadSpline class on simulated data. Test predictions (one step ahead forecasts).

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.quadSplineTest

def quadSplineTest2(): Unit

The quadSplineTest2 main function is used to test the QuadSpline class. Forecasting lake levels.

The quadSplineTest2 main function is used to test the QuadSpline class. Forecasting lake levels.

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.quadSplineTest2

def quadSplineTest3(): Unit

The quadSplineTest3 main function tests the QuadSpline class on real data: Forecasting lake levels. Test forecasts (1 to h steps ahead forecasts).

The quadSplineTest3 main function tests the QuadSpline class on real data: Forecasting lake levels. Test forecasts (1 to h steps ahead forecasts).

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.quadSplineTest3

def quadSplineTest4(): Unit

The quadSplineTest4 main function is used to test the QuadSpline class. Forecasting Fibonacci numbers.

The quadSplineTest4 main function is used to test the QuadSpline class. Forecasting Fibonacci numbers.

runMain scalation.modeling.forecasting.quadSplineTest4

Attributes

def randomWalkTest(): Unit

The randomWalkTest main function tests the RandomWalk class on simulated data. Test predictions (one step ahead forecasts).

The randomWalkTest main function tests the RandomWalk class on simulated data. Test predictions (one step ahead forecasts).

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.aRTest

def randomWalkTest2(): Unit

The randomWalkTest2 main function tests the RandomWalk2 class on real data: Forecasting lake levels. It compares with the other baselines: NullModel and TrendModel. Test predictions (one step ahead forecasts).

The randomWalkTest2 main function tests the RandomWalk2 class on real data: Forecasting lake levels. It compares with the other baselines: NullModel and TrendModel. Test predictions (one step ahead forecasts).

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.randomWalkTest2

def randomWalkTest3(): Unit

The randomWalkTest3 main function tests the RandomWalk class on real data: Forecasting lake levels. Test forecasts (1 to h steps ahead forecasts).

The randomWalkTest3 main function tests the RandomWalk class on real data: Forecasting lake levels. Test forecasts (1 to h steps ahead forecasts).

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.randomWalkTest3

def randomWalkTest4(): Unit

The randomWalkTest4 main function tests the RandomWalk class on real data: Forecasting COVID-19. Test forecasts (1 to h steps ahead forecasts).

The randomWalkTest4 main function tests the RandomWalk class on real data: Forecasting COVID-19. Test forecasts (1 to h steps ahead forecasts).

runMain scalation.modeling.forecasting.randomWalkTest4

Attributes

def rdot(b: VectorD, y_: VectorD, t: Int): Double

Compute a reverse dot product of the parameter vector b and the most recent actual values in the time series y_, going backwards from y_t. Use max (0, ..) to avoid using negative indices into the y_ vector.

Compute a reverse dot product of the parameter vector b and the most recent actual values in the time series y_, going backwards from y_t. Use max (0, ..) to avoid using negative indices into the y_ vector.

Value parameters

b

the parameter/coefficient vector (e.g., φ for AR)

t

the time point FROM WHICH to make prediction

y_

the actual values to use in making predictions

Attributes

def rdot(b: VectorD, yf: MatrixD, r: Int, c: Int): Double

Compute a reverse dot product of the parameter vector b and a diagonal of the the yf matrix starting at element (r, c) and moving up and back. Use max (0, ..) to avoid using negative indices into the yf matrix.

Compute a reverse dot product of the parameter vector b and a diagonal of the the yf matrix starting at element (r, c) and moving up and back. Use max (0, ..) to avoid using negative indices into the yf matrix.

Value parameters

b

the parameter/coefficient vector (e.g., φ for AR)

c

the starting column in the forecasting matrix (horizon)

r

the starting row in the forecasting matrix (time)

yf

the forecasting matrix (time x horizons)

Attributes

def rollingValidationTest(): Unit

The rollingValidationTest main function is used to test the rollValidate method in the RollingValidation object.

The rollingValidationTest main function is used to test the rollValidate method in the RollingValidation object.

runMain scalation.modeling.forecasting.rollingValidationTest

Attributes

def rollingValidationTest2(): Unit

The rollingValidationTest2 main function is used to test the rollValidate method in the RollingValidation object.

The rollingValidationTest2 main function is used to test the rollValidate method in the RollingValidation object.

runMain scalation.modeling.forecasting.rollingValidationTest2

Attributes

def rollingValidationTest3(): Unit

The rollingValidationTest3 main function is used to test the rollValidate method in the RollingValidation object. Compares baseline models on in-sample and out-of-sample assessment.

The rollingValidationTest3 main function is used to test the rollValidate method in the RollingValidation object. Compares baseline models on in-sample and out-of-sample assessment.

runMain scalation.modeling.forecasting.rollingValidationTest3

Attributes

def rollingValidationTest4(): Unit

The rollingValidationTest4 main function is used to test the rollValidate method in the RollingValidation object. Random Walk is used to make structure of the yf matrix clear.

The rollingValidationTest4 main function is used to test the rollValidate method in the RollingValidation object. Random Walk is used to make structure of the yf matrix clear.

runMain scalation.modeling.forecasting.rollingValidationTest4

Attributes

def sARIMATest(): Unit

The sARIMATest main function is used to test the SARIMA class.

The sARIMATest main function is used to test the SARIMA class.

runMain scalation.modeling.forecasting.sARIMATest

Attributes

def sARIMATest2(): Unit

The sARIMATest2 main function is used to test the SARIMA class.

The sARIMATest2 main function is used to test the SARIMA class.

runMain scalation.modeling.forecasting.sARIMATest2

Attributes

def sARIMATest3(): Unit

The sARIMATest3 main function is used to test the SARIMA class. Forecasting lake levels.

The sARIMATest3 main function is used to test the SARIMA class. Forecasting lake levels.

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.sARIMATest3

def sARIMATest4(): Unit

The sARIMATest4 main function is used to test the SARIMA class.

The sARIMATest4 main function is used to test the SARIMA class.

runMain scalation.modeling.forecasting.sARIMATest4

Attributes

def sARIMATest5(): Unit

The sARIMATest5 main function tests the SARIMA class on real data: Forecasting COVID-19.

The sARIMATest5 main function tests the SARIMA class on real data: Forecasting COVID-19.

runMain scalation.modeling.forecasting.sARIMATest5

Attributes

def sARIMATest6(): Unit

The sARIMATest6 main function is used to test the SARIMA class.

The sARIMATest6 main function is used to test the SARIMA class.

runMain scalation.modeling.forecasting.sARIMATest6

Attributes

def sARIMAXTest(): Unit

The sARIMAXTest main function is used to test the SARIMAX class.

The sARIMAXTest main function is used to test the SARIMAX class.

runMain scalation.modeling.forecasting.sARIMAXTest

Attributes

def sARIMAXTest2(): Unit

The sARIMAXTest2 main function is used to test the SARIMAX class.

The sARIMAXTest2 main function is used to test the SARIMAX class.

runMain scalation.modeling.forecasting.sARIMAXTest2

Attributes

def sARIMAXTest3(): Unit

The sARIMAXTest3 main function is used to test the SARIMAX class. Forecasting lake levels.

The sARIMAXTest3 main function is used to test the SARIMAX class. Forecasting lake levels.

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.sARIMAXTest3

def sARIMAXTest4(): Unit

The sARIMAXTest4 main function is used to test the SARIMAX class.

The sARIMAXTest4 main function is used to test the SARIMAX class.

runMain scalation.modeling.forecasting.sARIMAXTest4

Attributes

def sARIMAXTest5(): Unit

The sARIMAXTest5 main function tests the SARIMAX class on real data: Forecasting COVID-19.

The sARIMAXTest5 main function tests the SARIMAX class on real data: Forecasting COVID-19.

runMain scalation.modeling.forecasting.sARIMAXTest5

Attributes

def sARIMAXTest6(): Unit

The sARIMAXTest6 main function is used to test the SARIMAX class.

The sARIMAXTest6 main function is used to test the SARIMAX class.

runMain scalation.modeling.forecasting.sARIMAXTest6

Attributes

def simpleExpSmoothingTest(): Unit

The simpleExpSmoothingTest main function tests the AR class on simulated data. Test predictions (one step ahead forecasts).

The simpleExpSmoothingTest main function tests the AR class on simulated data. Test predictions (one step ahead forecasts).

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.simpleExpSmoothingTest

def simpleExpSmoothingTest2(): Unit

The simpleExpSmoothingTest2 main function is used to test the SimpleExpSmoothing class. Forecasting lake levels. Compare AR(1) and SimpleExpSmoothing models for the time series data.

The simpleExpSmoothingTest2 main function is used to test the SimpleExpSmoothing class. Forecasting lake levels. Compare AR(1) and SimpleExpSmoothing models for the time series data.

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.simpleExpSmoothingTest2

def simpleExpSmoothingTest3(): Unit

The simpleExpSmoothingTest3 main function is used to test the SimpleExpSmoothing class. Test customized smoothing (call smooth) versus optimized smoothing (call train).

The simpleExpSmoothingTest3 main function is used to test the SimpleExpSmoothing class. Test customized smoothing (call smooth) versus optimized smoothing (call train).

runMain scalation.modeling.forecasting.simpleExpSmoothingTest3

Attributes

def simpleExpSmoothingTest4(): Unit

The simpleExpSmoothingTest4 main function is used to test the SimpleExpSmoothing class. Test rolling validation.

The simpleExpSmoothingTest4 main function is used to test the SimpleExpSmoothing class. Test rolling validation.

runMain scalation.modeling.forecasting.simpleExpSmoothingTest4

Attributes

def simpleExpSmoothingTest5(): Unit

The simpleExpSmoothingTest5 main function is used to test the SimpleExpSmoothing class. Forecasting lake levels for several values of the smoothing parameter α.

The simpleExpSmoothingTest5 main function is used to test the SimpleExpSmoothing class. Forecasting lake levels for several values of the smoothing parameter α.

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.simpleExpSmoothingTest5

def simpleMovingAverageTest(): Unit

The simpleMovingAverageTest main function tests the SimpleMovingAverage class on simulated data.

The simpleMovingAverageTest main function tests the SimpleMovingAverage class on simulated data.

runMain scalation.modeling.forecasting.simpleMovingAverageTest

Attributes

def simpleMovingAverageTest2(): Unit

The simpleMovingAverageTest2 main function is used to test the SimpleMovingAverage class.

The simpleMovingAverageTest2 main function is used to test the SimpleMovingAverage class.

runMain scalation.modeling.forecasting.simpleMovingAverageTest2

Attributes

def simpleMovingAverageTest3(): Unit

The simpleMovingAverageTest3 main function is used to test the SimpleMovingAverage class. Forecasting lake levels.

The simpleMovingAverageTest3 main function is used to test the SimpleMovingAverage class. Forecasting lake levels.

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.simpleMovingAverageTest3

def simpleMovingAverageTest4(): Unit

The simpleMovingAverageTest4 main function is used to test the SimpleMovingAverage class. Decompose the lake levels dataset.

The simpleMovingAverageTest4 main function is used to test the SimpleMovingAverage class. Decompose the lake levels dataset.

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.simpleMovingAverageTest4

def simpleMovingAverageTest5(): Unit

The simpleMovingAverageTest5 main function is used to test the SimpleMovingAverage class.

The simpleMovingAverageTest5 main function is used to test the SimpleMovingAverage class.

runMain scalation.modeling.forecasting.simpleMovingAverageTest5

Attributes

def ssef(y: VectorD, yp: VectorD, skip: Int): Double

Compute the sum of squares errors assuming the first 'skip' error is zero.

Compute the sum of squares errors assuming the first 'skip' error is zero.

Value parameters

skip

skip this many elements at the beginning (defaults to 1)

y

the actual response vector

yp

the predicted response vector (one-step ahead)

Attributes

def stationarity_KPSSTest(): Unit

The stationarity_KPSSTest main function is used to test the Stationarity_KPSS class. Test whether white noise time-series has a unit root.

The stationarity_KPSSTest main function is used to test the Stationarity_KPSS class. Test whether white noise time-series has a unit root.

runMain scalation.modeling.forecasting.stationarity_KPSSTest

Attributes

def stationarity_KPSSTest2(): Unit

The stationarity_KPSSTest2 main function is used to test the Stationarity_KPSS class. Test whether the Lake Levels time-series has a unit root.

The stationarity_KPSSTest2 main function is used to test the Stationarity_KPSS class. Test whether the Lake Levels time-series has a unit root.

runMain scalation.modeling.forecasting.stationarity_KPSSTest2

Attributes

def stationarity_KPSSTest3(): Unit

The stationarity_KPSSTest3 main function is used to test the Stationarity_KPSS class. Test whether the differenced Lake Levels time-series has a unit root.

The stationarity_KPSSTest3 main function is used to test the Stationarity_KPSS class. Test whether the differenced Lake Levels time-series has a unit root.

runMain scalation.modeling.forecasting.stationarity_KPSSTest3

Attributes

def stationaryTest(): Unit

The stationaryTest main function tests the Stationary class on a simulated time-series.

The stationaryTest main function tests the Stationary class on a simulated time-series.

runMain scalation.modeling.forecasting.stationaryTest

Attributes

def stationaryTest2(): Unit

The stationaryTest2 main function tests the Stationary class on a simulated stationary time-series. An AR(1) is a stationary process when |φ_1| < 1, a unit root process when |φ_1| = 1, and explosive otherwise.

The stationaryTest2 main function tests the Stationary class on a simulated stationary time-series. An AR(1) is a stationary process when |φ_1| < 1, a unit root process when |φ_1| = 1, and explosive otherwise.

runMain scalation.modeling.forecasting.stationaryTest2

Attributes

def stationaryTest3(): Unit

The stationaryTest3 main function tests the Stationary class on a simulated stationary time-series. An AR(2) is a stationary process when |φ_2| < 1 and |φ_1| < 1 - φ_2, a unit root process when |φ_2| = 1 or |φ_1| = 1 - φ_2, and explosive otherwise.

The stationaryTest3 main function tests the Stationary class on a simulated stationary time-series. An AR(2) is a stationary process when |φ_2| < 1 and |φ_1| < 1 - φ_2, a unit root process when |φ_2| = 1 or |φ_1| = 1 - φ_2, and explosive otherwise.

runMain scalation.modeling.forecasting.stationaryTest3

Attributes

def testForecast(mod: Fit, y: VectorD, yf: MatrixD, p: Int): MatrixD

Test the actual response vector vs. forecasted matrix, returning the QoF for all forecasting horizons 1 to h. FIX - not agreeing with ForecasterUtil.testHorizons

Test the actual response vector vs. forecasted matrix, returning the QoF for all forecasting horizons 1 to h. FIX - not agreeing with ForecasterUtil.testHorizons

Value parameters

mod

the fittable model (one that extends Fit)

p

the number of variables/lags used in the model

y

the original actual response vector

yf

the forecasted response matrix

Attributes

def trendModelTest(): Unit

The trendModelTest main function tests the TrendModel class on simulated data. Test predictions (one step ahead forecasts).

The trendModelTest main function tests the TrendModel class on simulated data. Test predictions (one step ahead forecasts).

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.trendModelTest

def trendModelTest2(): Unit

The trendModelTest2 main function tests the TrendModel class on real data: Forecasting lake levels. Test predictions (one step ahead forecasts).

The trendModelTest2 main function tests the TrendModel class on real data: Forecasting lake levels. Test predictions (one step ahead forecasts).

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.trendModelTest2

def trendModelTest3(): Unit

The trendModelTest3 main function tests the TrendModel class on real data: Forecasting lake levels. Test forecasts (1 to h steps ahead forecasts).

The trendModelTest3 main function tests the TrendModel class on real data: Forecasting lake levels. Test forecasts (1 to h steps ahead forecasts).

Attributes

See also

cran.r-project.org/web/packages/fpp/fpp.pdf

runMain scalation.modeling.forecasting.trendModelTest3

def undel(v: VectorD, y0: Double): VectorD

Return the undifferenced time-series from the velocity series.

Return the undifferenced time-series from the velocity series.

Value parameters

v

the differenced time-series (velocity)

y0

the first value in the original time-series

Attributes

def varTest(): Unit

The varTest main function tests the VAR class. This test is used to CHECK that the makeExoCols method (@see apply) is working correctly. May get NaN for some maximum lags (p) due to multi-collinearity.

The varTest main function tests the VAR class. This test is used to CHECK that the makeExoCols method (@see apply) is working correctly. May get NaN for some maximum lags (p) due to multi-collinearity.

runMain scalation.modeling.forecasting.varTest

Attributes

def varTest2(): Unit

The varTest2 main function tests the VAR class on real data: Forecasting Gas Furnace Data. Performs In-Sample Testing.

The varTest2 main function tests the VAR class on real data: Forecasting Gas Furnace Data. Performs In-Sample Testing.

runMain scalation.modeling.forecasting.varTest2

Attributes

def varTest3(): Unit

The varTest3 main function tests the VAR class on real data: Forecasting COVID-19 Weekly Data. Performs In-Sample Testing. Goal: Find the variable that works best with "new_deaths"

The varTest3 main function tests the VAR class on real data: Forecasting COVID-19 Weekly Data. Performs In-Sample Testing. Goal: Find the variable that works best with "new_deaths"

runMain scalation.modeling.forecasting.varTest3

Attributes

def varTest4(): Unit

The varTest4 main function tests the VAR class on real data: Forecasting COVID-19 Weekly Data. Performs In-Sample Testing. Goal: Find the four variables that works best with "new_deaths"

The varTest4 main function tests the VAR class on real data: Forecasting COVID-19 Weekly Data. Performs In-Sample Testing. Goal: Find the four variables that works best with "new_deaths"

runMain scalation.modeling.forecasting.varTest4

Attributes

def varTest5(): Unit

The varTest5 main function tests the VAR class on real data: Forecasting COVID-19 Weekly Data. Does TnT Testing on endogenous and exogenous variables. Determine the terms to include in the model using Stepwise on In-Sample.

The varTest5 main function tests the VAR class on real data: Forecasting COVID-19 Weekly Data. Does TnT Testing on endogenous and exogenous variables. Determine the terms to include in the model using Stepwise on In-Sample.

runMain scalation.modeling.forecasting.varTest5

Attributes

def varTest6(): Unit

The varTest6 main function tests the VAR class on real data: Forecasting COVID-19 Weekly Data. Does Rolling Validation on variables. Determine the terms to include in the model using Stepwise on In-Sample.

The varTest6 main function tests the VAR class on real data: Forecasting COVID-19 Weekly Data. Does Rolling Validation on variables. Determine the terms to include in the model using Stepwise on In-Sample.

runMain scalation.modeling.forecasting.varTest6

Attributes

inline def Δ(y: VectorD): VectorD
Generated with