added pkgs to various notes; geo-sptemp >> notes on args for EstiStAni; quarto-docs >> added slides sect

Author: ercbk

qmd/association-time-series.qmd

@@ -125,6 +125,9 @@
 -   The smaller the total distance between two series, the greater the association.
 -   Distance algorithms in the [{dtw}]{style="color: #990000"} and [{dtwclust}]{style="color: #990000"} packages
 -   Other packages
+    -   [{]{style="color: #990000"}[dCovTS](https://cran.r-project.org/web/packages/dCovTS/index.html){style="color: #990000"}[}]{style="color: #990000"} - Distance Covariance and Correlation for Time Series Analysis
+        -   Both versions of biased and unbiased estimators of distance covariance and correlation are provided.
+        -   Test statistics for testing pairwise independence are also implemented. Some data sets are also included.
     -   [{]{style="color: #990000"}[distantia](https://github.com/BlasBenito/distantia){style="color: #990000"}[}]{style="color: #990000"} - Qantifies dissimilarity between time series
         -   Types of ts
             -   Multivariate and univariate time series.

qmd/feature-engineering-time-series.qmd

@@ -666,6 +666,7 @@
         -   Uses a detector statistic that is carefully devised to detect changes in the tail pairwise dependence matrix (TPDM)
         -   The TPDM acts analogously to the covariance matrix in classical Gaussian statistics, which has made it an increasingly popular tool in the study of multivariate extremes
         -   Example analyzes multivariate EEG signals from seizure-prone neonatal subjects
+    -   [{]{style="color: #990000"}[rbreak](https://cran.r-project.org/web/packages/rbreak/index.html){style="color: #990000"}[}]{style="color: #990000"} - Methods for detecting structural breaks and estimating break locations for linear multiple regression models under general linear restrictions on the coefficient vector.
     -   [{]{style="color: #990000"}[RegimeChange](https://cran.r-project.org/web/packages/RegimeChange/index.html){style="color: #990000"}[}]{style="color: #990000"} - A unified framework for detecting regime changes (changepoints) in time series data.
         -   Implements both frequentist methods including Cumulative Sum (CumSum), Pruned Exact Linear Time (PELT), Binary Segmentation, and Wild Binary Segmentation, as well as Bayesian methods such as Bayesian Online Changepoint Detection (BOCPD), Shiryaev-Roberts
     -   [{]{style="color: #990000"}[robcp](https://cran.r-project.org/web/packages/robcp/index.html){style="color: #990000"}[}]{style="color: #990000"} - Robust Change-Point Tests

qmd/forecasting-multivariate.qmd

@@ -4,22 +4,32 @@
 
 -   Multivariate time series analysis seeks to analyze several time series jointly. The rationale behind this is the possible presence of interdependences between the different time series.
 -   Packages
-    -   [{]{style="color: #990000"}[bayesianVARs](https://luisgruber.github.io/bayesianVARs/){style="color: #990000"}[}]{style="color: #990000"} ([Papers](https://cran.r-project.org/web/packages/bayesianVARs/index.html)) - Efficient Markov Chain Monte Carlo (MCMC) algorithms for the fully **Bayesian** estimation of vectorautoregressions (VARs) featuring **stochastic volatility** (SV)
-    -   [{]{style="color: #990000"}[cointsmall](https://cran.r-project.org/web/packages/cointsmall/index.html){style="color: #990000"}[}]{style="color: #990000"} - **Cointegration Tests** with Structural Breaks in Small Samples
-    -   [{]{style="color: #990000"}[fbardl](https://cran.r-project.org/web/packages/fbardl/index.html){style="color: #990000"}[}]{style="color: #990000"} - Fourier Bootstrap ARDL **Cointegration Test**
-    -   [{]{style="color: #990000"}[hatemicoint](https://cran.r-project.org/web/packages/hatemicoint/index.html){style="color: #990000"}[}]{style="color: #990000"} - Hatemi-J **Cointegration Test** with Two Unknown Regime Shifts
-    -   [{]{style="color: #990000"}[l1rotation](https://cran.r-project.org/web/packages/l1rotation/index.html){style="color: #990000"}[}]{style="color: #990000"} - **Identify Loading Vectors** under Sparsity in Factor Models
-        -   Uses a sparse representation to overcome rotational indeterminacy and to be able find the true interpretation of how the factors and variables are related.
-    -   [{]{style="color: #990000"}[makicoint](https://cran.r-project.org/web/packages/makicoint/index.html){style="color: #990000"}[}]{style="color: #990000"} - Maki **Cointegration Test** with Structural Breaks
-    -   [{]{style="color: #990000"}[micvar](https://cran.r-project.org/web/packages/micvar/index.html){style="color: #990000"}[}]{style="color: #990000"} - Implements order selection for Vector Autoregressive (VAR) models using the Mean Square Information Criterion (MIC).
-        -   Unlike standard methods such as AIC and BIC, MIC is likelihood-free. This method consistently estimates VAR order and has robust performance under model misspecification
-    -   [{]{style="color: #990000"}[mvardlurt](https://cran.r-project.org/web/packages/mvardlurt/index.html){style="color: #990000"}[}]{style="color: #990000"} - Multivariate **ARDL Unit Root Test**
-    -   [{]{style="color: #990000"}[mvgam](https://nicholasjclark.github.io/mvgam/){style="color: #990000"}[}]{style="color: #990000"} - **Bayesian (Dynamic) GAMs**
-        -   Constructs State-Space models that can include highly flexible nonlinear predictor effects for both process and observation components by leveraging functionalities from the impressive brms and mgcv packages
-    -   [{]{style="color: #990000"}[NetVAR](https://cran.r-project.org/web/packages/NetVAR/index.html){style="color: #990000"}[}]{style="color: #990000"} - VAR Models with tailored regularization structures are provided to **uncover network type structures** in the data, such as influential time series (**influencers**).
-    -   [{]{style="color: #990000"}[nonParQuantileCausality](https://cran.r-project.org/web/packages/nonParQuantileCausality/index.html){style="color: #990000"}[}]{style="color: #990000"} - Nonparametric Causality in Quantiles Test (granger causality type of test)
-    -   [{]{style="color: #990000"}[statioVAR](https://cran.r-project.org/web/packages/statioVAR/index.html){style="color: #990000"}[}]{style="color: #990000"} - **Trend Removal** for Vector Autoregressive Workflows
-    -   [{]{style="color: #990000"}[VARcpDetectOnline](https://cran.r-project.org/web/packages/VARcpDetectOnline/index.html){style="color: #990000"}[}]{style="color: #990000"} - Sequential Change Point Detection for High-Dimensional VAR Models. Effectively **identifies shifts in temporal and cross-correlations** within high-dimensional time series data
+    -   [Cointegration and Granger Causality]{.underline}
+        -   [{]{style="color: #990000"}[caustests](https://cran.r-project.org/web/packages/caustests/index.html){style="color: #990000"}[}]{style="color: #990000"} - Multiple **Granger Causality** Tests
+        -   [{]{style="color: #990000"}[cointsmall](https://cran.r-project.org/web/packages/cointsmall/index.html){style="color: #990000"}[}]{style="color: #990000"} - Cointegration Tests with **Structural Breaks in Small Samples**
+        -   [{]{style="color: #990000"}[cointests](https://cran.r-project.org/web/packages/cointests/index.html){style="color: #990000"}[}]{style="color: #990000"} - Comprehensive Cointegration Tests with **Fourier and Panel Methods**
+            -   Fourier-based cointegration tests (FADL, FEG, FEG2, Tsong) that accommodate smooth structural breaks via flexible Fourier terms
+            -   Panel CADF cointegration tests with structural breaks using the Common Correlated Effects (CCE) estimator
+        -   [{]{style="color: #990000"}[fbardl](https://cran.r-project.org/web/packages/fbardl/index.html){style="color: #990000"}[}]{style="color: #990000"} - **Fourier Bootstrap ARDL** Cointegration Test
+        -   [{]{style="color: #990000"}[hatemicoint](https://cran.r-project.org/web/packages/hatemicoint/index.html){style="color: #990000"}[}]{style="color: #990000"} - **Hatemi-J** Cointegration Test with Two Unknown Regime Shifts
+        -   [{]{style="color: #990000"}[makicoint](https://cran.r-project.org/web/packages/makicoint/index.html){style="color: #990000"}[}]{style="color: #990000"} - **Maki** Cointegration Test with Structural Breaks
+        -   [{]{style="color: #990000"}[nonParQuantileCausality](https://cran.r-project.org/web/packages/nonParQuantileCausality/index.html){style="color: #990000"}[}]{style="color: #990000"} - Nonparametric Causality in Quantiles Test (granger causality type of test)
+        -   [{]{style="color: #990000"}[xtdhcoint](https://cran.r-project.org/web/packages/xtdhcoint/index.html){style="color: #990000"}[}]{style="color: #990000"} - **Durbin-Hausman Panel** Cointegration Tests
+            -   Robust to cross-sectional dependence through common factor extraction using principal components.
+            -   Provides both group-mean (DHg) and panel (DHp) test statistics with automatic factor number selection via information criteria
+    -   [Vector Autoregression (VAR)]{.underline}
+        -   [{]{style="color: #990000"}[bayesianVARs](https://luisgruber.github.io/bayesianVARs/){style="color: #990000"}[}]{style="color: #990000"} ([Papers](https://cran.r-project.org/web/packages/bayesianVARs/index.html)) - Efficient Markov Chain Monte Carlo (MCMC) algorithms for the fully **Bayesian** estimation of vectorautoregressions (VARs) featuring **stochastic volatility** (SV)
+        -   [{]{style="color: #990000"}[micvar](https://cran.r-project.org/web/packages/micvar/index.html){style="color: #990000"}[}]{style="color: #990000"} - Implements **order selection for Vector Autoregressive** (VAR) models using the Mean Square Information Criterion (MIC).
+            -   Unlike standard methods such as AIC and BIC, MIC is likelihood-free. This method consistently estimates VAR order and has robust performance under model misspecification
+        -   [{]{style="color: #990000"}[NetVAR](https://cran.r-project.org/web/packages/NetVAR/index.html){style="color: #990000"}[}]{style="color: #990000"} - VAR Models with tailored regularization structures are provided to **uncover network type structures** in the data, such as influential time series (**influencers**).
+        -   [{]{style="color: #990000"}[statioVAR](https://cran.r-project.org/web/packages/statioVAR/index.html){style="color: #990000"}[}]{style="color: #990000"} - **Trend Removal** for Vector Autoregressive Workflows
+        -   [{]{style="color: #990000"}[VARcpDetectOnline](https://cran.r-project.org/web/packages/VARcpDetectOnline/index.html){style="color: #990000"}[}]{style="color: #990000"} - Sequential Change Point Detection for High-Dimensional VAR Models. Effectively **identifies shifts in temporal and cross-correlations** within high-dimensional time series data
+    -   [General]{.underline}
+        -   [{]{style="color: #990000"}[l1rotation](https://cran.r-project.org/web/packages/l1rotation/index.html){style="color: #990000"}[}]{style="color: #990000"} - **Identify Loading Vectors** under Sparsity in Factor Models
+            -   Uses a sparse representation to overcome rotational indeterminacy and to be able find the true interpretation of how the factors and variables are related.
+        -   [{]{style="color: #990000"}[mvardlurt](https://cran.r-project.org/web/packages/mvardlurt/index.html){style="color: #990000"}[}]{style="color: #990000"} - Multivariate **ARDL Unit Root Test**
+        -   [{]{style="color: #990000"}[mvgam](https://nicholasjclark.github.io/mvgam/){style="color: #990000"}[}]{style="color: #990000"} - **Bayesian (Dynamic) GAMs**
+            -   Constructs State-Space models that can include highly flexible nonlinear predictor effects for both process and observation components by leveraging functionalities from the impressive brms and mgcv packages
 -   Resources
     -   [Model Checking for Vector Autoregressive Models](https://osf.io/preprints/psyarxiv/k6uz4_v2) ([Github](https://github.com/jmbh/ModelCheckingForVAR)) - Diagnostics for VAR models
 -   Papers

qmd/forecasting-statistical.qmd

@@ -186,6 +186,10 @@
         -   KPSS-type tests are intended to complement unit root tests, such as the Dickey–Fuller tests. By testing both the unit root hypothesis and the stationarity hypothesis, one can distinguish series that appear to be stationary, series that appear to have a unit root, and series for which the data (or the tests) are not sufficiently informative to be sure whether they are stationary.
         -   [{]{style="color: #990000"}[boundedur](https://cran.r-project.org/web/packages/boundedur/index.html){style="color: #990000"}[}]{style="color: #990000"} - Unit Root Tests for Bounded Time Series
         -   [{]{style="color: #990000"}[qadf](https://cran.r-project.org/web/packages/qadf/index.html){style="color: #990000"}[}]{style="color: #990000"} - Quantile Autoregressive Distributed Lag Unit Root Test
+        -   [unitrootests](https://cran.r-project.org/web/packages/unitrootests/index.html) - Comprehensive Unit Root and Stationarity Tests
+            -   Quantile ADF tests ([Paper](https://doi.org/10.1198%2F016214504000001114))
+            -   GARCH-based unit root tests with endogenous structural breaks ([Paper](https://doi.org/10.1016%2Fj.eneco.2014.11.021))
+            -   Comprehensive Dickey-Fuller, Phillips-Perron, KPSS, ERS/DF-GLS, Zivot-Andrews, and Kobayashi-McAleer tests with an Elder-Kennedy decision strategy ([Paper](https://doi.org/10.1080%2F00220480109595179))
     -   Steps:
         1.  ADF:
             1.  If H~0~ rejected. The trend (if any) can be represented by a deterministic linear trend.
@@ -334,7 +338,7 @@
 -   Drift
 
     | **Feature** | **Random Walk without Drift** | **Random Walk with Drift** |
-    |:---|:---|:---|
+    |:-----------------------|:-----------------------|:-----------------------|
     | Steps | Purely random, equal probability left/right | Biased, one direction slightly more likely |
     | Change in value | Average change is zero | Average change includes a constant drift |
     | Path | Zig-zag around starting point | Zig-zag with upward/downward trend |

qmd/geospatial-spat-temp.qmd

@@ -24,12 +24,14 @@
         -   Highly parallelizable and hence, much faster than traditional Markov chain Monte Carlo algorithms while delivering competitive predictive performance
         -   Fits **Bayesian spatially-temporally varying coefficients (STVC) generalized linear models** without MCMC
     -   [{]{style="color: #990000"}[stdbscan](https://cran.r-project.org/web/packages/stdbscan/index.html){style="color: #990000"}[}]{style="color: #990000"} - Spatio-Temporal **DBSCAN Clustering**
+    -   [{]{style="color: #990000"}[spacetime](https://cran.r-project.org/web/packages/spacetime/index.html){style="color: #990000"}[}]{style="color: #990000"} - Superceded by [{stars}]{style="color: #990000"}. Extends the shared classes defined in [{sp}]{style="color: #990000"} for spatio-temporal data.
+        -   [{stars}]{style="color: #990000"} uses PROJ and GDAL through [{sf}]{style="color: #990000"}.
     -   [{]{style="color: #990000"}[spTimer](https://cran.r-project.org/web/packages/spTimer/index.html){style="color: #990000"}[}]{style="color: #990000"} ([Vignette](https://www.jstatsoft.org/article/view/v063i15)) - Spatio-Temporal **Bayesian Modelling**
         -   Models: Bayesian Gaussian Process (GP) Models, Bayesian Auto-Regressive (AR) Models, and Bayesian Gaussian Predictive Processes (GPP) based AR Models for spatio-temporal big-n problems
         -   Depends on [{spacetime}]{style="color: #990000"} and [{sp}]{style="color: #990000"}
     -   [{]{style="color: #990000"}[stars](https://r-spatial.github.io/stars/){style="color: #990000"}[}]{style="color: #990000"} - **Reading, manipulating, writing and plotting** spatiotemporal arrays (raster and vector data cubes) in 'R', using 'GDAL' bindings provided by 'sf', and 'NetCDF' bindings by 'ncmeta' and 'RNetCDF'
         -   Only handles **full lattice/grid** data
-        -   It supercedes the [{]{style="color: #990000"}[spacetime](https://cran.r-project.org/web/packages/spacetime/index.html){style="color: #990000"}[}]{style="color: #990000"}, which extended the shared classes defined in [{sp}]{style="color: #990000"} for spatio-temporal data. [{stars}]{style="color: #990000"} uses PROJ and GDAL through [{sf}]{style="color: #990000"}.
+        -   It supercedes [{spacetime}]{style="color: #990000"}, which extended the shared classes defined in [{sp}]{style="color: #990000"} for spatio-temporal data. [{stars}]{style="color: #990000"} uses PROJ and GDAL through [{sf}]{style="color: #990000"}.
             -   Easily convert spacetime objects to stars object with `st_as_stars(spacetime_obj)`
         -   Has [dplyr verb methods](https://r-spatial.github.io/stars/reference/dplyr.html)
     -   [{]{style="color: #990000"}[WaveST](https://cran.r-project.org/web/packages/WaveST/index.html){style="color: #990000"}[}]{style="color: #990000"} - An integrated **wavelet-based spatial time series modeling** framework designed to enhance predictive accuracy under noisy and nonstationary conditions by jointly exploiting multi-resolution (wavelet) information and spatial dependence
@@ -1860,11 +1862,26 @@
     -   See [Terms](geospatial-spat-temp.qmd#sec-geo-sptemp-terms){style="color: green"} \>\> Anisotropy, [`vgmST`](geospatial-spat-temp.qmd#sec-geo-sptemp-krig-var-vgmst){style="color: green"} \>\> Metric, Sum Metric, and Simple Sum Metric
     -   Default is NA and will be understood as identity (1 temporal unit = 1 spatial unit)
         -   Docs says this assumption is rarely the case so could be worth including to see if it improves performance
+    -   Arguments
+        -   [empVgm]{.arg-text} - Empircal variogram (`variogramST`)
+        -   [interval]{.arg-text} - The interval to search for the stAni value
+        -   [method]{.arg-text} - Method used to estimate the value (see below)
+        -   [spatialVgm]{.arg-text} - `vgm` specification for the spatial model
+        -   [temporalVgm]{.arg-text} - `vgm` specification for the temporal model
+        -   [s.range]{.arg-text} - Spatial cutoff value (for linear model)
+        -   [t.range]{.arg-text} - Temporal cutoff value (for linear model)
     -   Anisotropy Estimation Heuristics using `estiStAni`
         -   [method = "linear"]{.arg-text} - Rescales a linear model
+            -   Requires [empVgm]{.arg-text}, [interval]{.arg-text} arguments
+            -   Recommended to also specify [s.range]{.arg-text} and [t.range]{.arg-text} arguments
+                -   i.e. where *before* this value, there's a sharp increase. The function will drop values after this distance/time in order to get a good linear fit.
+                -   Technically the function will run without these value set but you probably need to set them in most cases to get an unbiased estimate
         -   [method = "range"]{.arg-text} - Estimates equal ranges
+            -   Requires [empVgm]{.arg-text}, [spatialVgm]{.arg-text}, and [temporalVgm]{.arg-text} arguments
         -   [method = "vgm"]{.arg-text} - Rescales a pure spatial variogram
+            -   Requires [empVgm]{.arg-text}, [interval]{.arg-text}, and [spatialVgm]{.arg-text} arguments
         -   [method = "metric"]{.arg-text} - Estimates a complete spatio-temporal metric variogram model and returns its spatio-temporal anisotropy parameter.
+            -   Requires [empVgm]{.arg-text}, [interval]{.arg-text}, and [spatialVgm]{.arg-text} arguments
 
 #### Diagnostics {#sec-geo-sptemp-krig-var-diag .unnumbered}