Learn R Programming

generalCorr (version 1.2.6)

Generalized Correlations, Causal Paths and Portfolio Selection

Description

Function gmcmtx0() computes a more reliable (general) correlation matrix. Since causal paths from data are important for all sciences, the package provides many sophisticated functions. causeSummBlk() and causeSum2Blk() give easy-to-interpret causal paths. Let Z denote control variables and compare two flipped kernel regressions: X=f(Y, Z)+e1 and Y=g(X, Z)+e2. Our criterion Cr1 says that if |e1*Y|>|e2*X| then variation in X is more "exogenous or independent" than in Y, and the causal path is X to Y. Criterion Cr2 requires |e2|<|e1|. These inequalities between many absolute values are quantified by four orders of stochastic dominance. Our third criterion Cr3, for the causal path X to Y, requires new generalized partial correlations to satisfy |r*(x|y,z)|< |r*(y|x,z)|. The function parcorVec() reports generalized partials between the first variable and all others. The package provides several R functions including get0outliers() for outlier detection, bigfp() for numerical integration by the trapezoidal rule, stochdom2() for stochastic dominance, pillar3D() for 3D charts, canonRho() for generalized canonical correlations, depMeas() measures nonlinear dependence, and causeSummary(mtx) reports summary of causal paths among matrix columns. Portfolio selection: decileVote(), momentVote(), dif4mtx(), exactSdMtx() can rank several stocks. Functions whose names begin with 'boot' provide bootstrap statistical inference, including a new bootGcRsq() test for "Granger-causality" allowing nonlinear relations. A new tool for evaluation of out-of-sample portfolio performance is outOFsamp(). Panel data implementation is now included. See eight vignettes of the package for theory, examples, and usage tips. See Vinod (2019) \doi{10.1080/03610918.2015.1122048}.

Copy Link

Version

Install

install.packages('generalCorr')

Monthly Downloads

478

Version

1.2.6

License

GPL (>= 2)

Maintainer

Hrishikesh Vinod

Last Published

October 9th, 2023

Functions in generalCorr (1.2.6)

Absolute values of gradients (apd's) of kernel regressions of x on y when both x and y are standardized and control variables are present.

Compute matrix of n999 rows and p-1 columns of bootstrap `sum' (scores from Cr1 to Cr3).

Compute the numerical integration by the trapezoidal rule.

Probability of unambiguously correct (+ or -) sign from bootPairs output

Probability of unambiguously correct (+ or -) sign from bootPairs output transformed to percentages.

bootstrap confidence intervals for (x2-x1) exact SD1 to SD4 stochastic dominance .

Absolute residuals kernel regressions of standardized x on y and control variables, Cr1 has abs(RHS*y) not gradients.

Absolute values of Hausman-Wu null in kernel regressions of x on y when both x and y are standardized.

Compute matrix of n999 rows and p-1 columns of bootstrap `sum' (strength from Cr1 to Cr3).

Compute vector of n999 nonlinear Granger causality paths

Compute vector of n999 nonlinear Granger causality paths

Block Version 2: Kernel causality summary of causal paths from three criteria

No print (NoP) version of causeSummBlk summary causal paths from three criteria

Kernel causality summary of evidence for causal paths from three criteria

Compute matrix of n999 rows and p-1 columns of bootstrap `sum' index (strength from older criterion Cr1, with newer Cr2 and Cr3).

Compute confidence intervals [quantile(s)] of indexes from bootPairs output

Older Kernel causality summary of evidence for causal paths from three criteria

All Pair Version Kernel (block) causality summary paths from three criteria

Generalized canonical correlation, estimating alpha, beta, rho.

internal da2Lag

Block Version 2: Kernel causality summary of causal paths from three criteria

Function compares nine deciles of stock return distributions.

Kernel regressions based causal paths in Panel Data.

causeSummary2NoP

No Print version Kernel causality summary of evidence for causal paths from three criteria using new exact stochastic dominance.

Kernel causality summary of evidence for causal paths from three criteria using new exact stochastic dominance. The function develops a unanimity index for deciding which flip (y on xi) or (xi on y) is best. Relevant signs determine the causal direction and unanimity index among three criteria. While allowing the researcher to keep some variables as controls, or outside the scope of causal path determination (e.g., age or latitude) this function produces detailed causal path information in a 5 column matrix identifying the names of variables, causal path directions, path strengths re-scaled to be in the range [--100, 100], (table reports absolute values of the strength) plus Pearson correlation and its p-value. The `2' in the name of the function suggests a second implementation where exact stochastic dominance, decileVote, and momentVote are used and where we avoid Anderson's trapezoidal approximation.

depMeas Signed measure of nonlinear nonparametric dependence between two vectors.

Compute usual summary stats of 'sum' indexes from bootPairs output

order 4 differencing of a time series vector

Compute usual summary stats of 'sum' index in (-100, 100) from bootPair2

Compute cofactor of a matrix based on row r and column c.

Function to compute outliers and their count using Tukey's method using 1.5 times interquartile range (IQR) to define boundaries.

Compares two vectors (portfolios) using momentVote, DecileVote and exactSdMtx functions.

Two sequences: starting+ending values from n and blocksize (internal use)

Compares two vectors (portfolios) using stochastic dominance of orders 1 to 4.

Matrix R* of generalized correlation coefficients captures nonlinearities.

Function to compute generalized correlation coefficients r*(x|y) and r*(y|x) from two vectors (not matrices)

compute the matrix R* of generalized correlation coefficients.

Heuristic t test of the difference between two generalized correlations.

internal goodCol

Exact stochastic dominance computation from areas above ECDF pillars.

generalCorrInfo

generalCorr package description:

internal object

order four differencing of a matrix of time series

Function compares Pearson Stats and Sharpe Ratio for a matrix of stock returns

Kernel regression with control variables and optional residuals and gradients. version 2 regtype="ll" for local linear, bwmethod="cv.aic" for AIC-based bandwidth selection. It admits control variables.

Kernel regression with control variables and optional residuals and gradients.

Internal diff.e0

Approximate overall magnitudes of kernel regression partials dx/dy and dy/dx.

After removing control variables, magnitude of effect of x on y, and of y on x.

Kernel regression version 2 with optional residuals and gradients with regtype="ll" for local linear, bwmethod="cv.aic" for AIC-based bandwidth selection.

Function to do matched deletion of missing rows from x, y and control variable(s).

internal nam.badCol

Kernel regression with options for residuals and gradients.

Function to do matched deletion of missing rows from x, y and z variable(s).

internal nam.goodCol

Matrix R* of generalized correlation coefficients captures nonlinearities using blocks.

Block version reports many generalized partial correlation coefficients allowing control variables.

Block version of generalized partial correlation coefficients between Xi and Xj, after removing the effect of xk, via nonparametric regression residuals.

Compare out-of-sample (short) selling algorithms by a leave-percent-out method.

Generalized partial correlation coefficients between Xi and Xj, after removing the effect of Xk, via OLS regression residuals.

internal nam.mtx0

internal min.e0

Function to do pairwise deletion of missing rows.

Function to do compute the minor of a matrix defined by row r and column c.

Generalized partial correlation coefficients between Xi and Xj,

Report many generalized partial correlation coefficients allowing control variables.

Matrix of generalized partial correlation coefficients, always leaving out control variables, if any.

Vector of hybrid generalized partial correlation coefficients.

Partial correlation coefficient between Xi and Xj after removing the linear effect of all others.

Compute generalized (ridge-adjusted) partial correlation coefficients from matrix R*. (deprecated)

Compare out-of-sample portfolio choice algorithms by a leave-percent-out method.

Vector of hybrid generalized partial correlation coefficients.

Generalized partial correlation coefficients between Xi and Xj, after removing the effect of xk, via nonparametric regression residuals.

Generalized partial correlation coefficient between Xi and Xj after removing the effect of all others. (older version, deprecated)

Vector of generalized partial correlation coefficients (GPCC), always leaving out control variables, if any.

Silently compute generalized (ridge-adjusted) partial correlation coefficients from matrix R*.

internal rijMrji

Compute the portfolio return knowing the rank of a stock in the input `mtx'.

Compute the portfolio return knowing the rank of a stock in the input `mtx'. This function computes the return earned knowing the rank of a stock computed elsewhere and named myrank associate with the data columns in the input mtx of stock returns. For example, mtx has p=28 Dow Jones stocks over n=169 monthly returns. Portfolio weights are assumed to be linearly declining. If maxChosen=4, the weights are 1/10, 2/10, 3/10 and 4/10, which add up to unity. These portfolio weights are assigned in their order in the sense that first chosen stock (choice rank =p) gets portfolio weight=4/10. The function computes return from the stocks using the `myrank' argument. This helps in assessing out-of-sample performance of (short) the strategy of selling lowest ranking stocks. It is mostly for internal use by outOFsell(). This is a sell version of rank2return().

Compute the bootstrap probability of correct causal direction.

Create a 3D pillar chart to display (x, y, z) data coordinate surface.

Intermediate weighting function giving Non-Expected Utility theory weights.

Compute probability of positive or negative sign from bootPairs output

internal rhs.lag2

internal salesLag

internal ridgek

Function to compute generalized correlation coefficients r*(x,y).

internal sales2Lag

internal sgn.e0

Block Version of silentPair2 for causality scores with control variables

Function reporting kernel causality results as a 7-column matrix.(deprecated)

Older kernel-causality unanimity score matrix with optional control variables

Block Version of silentPairs for causality scores with control variables

No-print kernel-causality unanimity score matrix with optional control variables

Function reporting detailed kernel causality results in a 7-column matrix (uses deprecated criterion 1, no longer recommended but may be useful for second and third criterion typ=2,3)

Kernel causality computations admitting control variables.

Summary magnitudes after removing control variables in several pairs where dependent variable is fixed.

Older version, kernel causality weighted sum allowing control variables

Kernel causality computations admitting control variables reporting a 7-column matrix, version 2.

No-print kernel causality scores with control variables Hausman-Wu Criterion 1

kernel causality (version 2) scores with control variables

Function reporting kernel causality results as a 7-column matrix, version 2.

internal sort.e0

internal sort.abse0

Pseudo regression coefficients from generalized partial correlation coefficients, (GPCC).

sudoCoefParcorH

Peudo regression coefficients from hybrid generalized partial correlation coefficients (HGPCC).

Compute ranks of rows of matrix and summarize them into a choice suggestion.

Creates input for the stochastic dominance function stochdom2

Replace asymmetric matrix by max of abs values of [i,j] or [j,i] elements.

Sort all columns of matrix x with respect to the j-th column.

Residuals of kernel regressions of x on y when both x and y are standardized.

Standardize x and y vectors to achieve zero mean and unit variance.

Compute vectors measuring stochastic dominance of four orders.

Nonlinear Granger causality between two time series workhorse function.

Generalized Granger-Causality. If dif>0, x2 Granger-causes x1.

Nonlinear Granger causality between two time series workhorse function.(local constant version)

European Crime Data

Generalized Granger-Causality. If dif>0, x2 Granger-causes x1.

Function to compute a vector of 2 lagged values of a variable from panel data.

Compute fitted values from kernel regression of x on y and y on x

Function for computing a vector of one-lagged values of xj, a variable from panel data.

Block version abs_stdrhser Absolute residuals kernel regressions of standardized x on y and control variables, Cr1 has abs(Resid*RHS).

Report causal identification for all pairs of variables in a matrix (deprecated function). It is better to choose a target variable and pair it with all others, instead of considering all possible targets.

internal badCol

Absolute values of gradients (apd's) of kernel regressions of x on y when both x and y are standardized.

Absolute residuals of kernel regression of x on y.

Block version of Absolute values of residuals of kernel regressions of standardized x on standardized y and control variables.

Block version of abs-stdres Absolute values of residuals of kernel regressions of standardized x on standardized y, no control variables.

Absolute values of residuals of kernel regressions of x on y when both x and y are standardized and control variables are present (C for control presence).

Absolute values of residuals of kernel regressions of x on y when both x and y are standardized.