SQL functions for linear regression. More...

Functions
aggregate robust_linregr_result	robust_linregr (float8 dependentVariable, float8[] independentVariables, float8[] coef)
	Compute robust regression diagnostic statistics for linear regression. More...

CREATE FUNCTION MADlib	__internal_get_robust_linregr_result (source_table VARCHAR--name of input table, dependent_varname VARCHAR--name of dependent variable, independent_varname VARCHAR--name of independent variable, linregr_coeffs DOUBLE PRECISION[]--coeffs from linear regression) RETURNS MADlib.robust_linregr_result AS $$DECLARE robust_value MADlib.robust_linregr_result
	Return robust linear regression estimates given a set of coefficients.

CREATE FUNCTION MADlib	__internal_get_robust_linregr_insert_string (robust_lin_rst MADlib.robust_linregr_result, linregr_coeffs DOUBLE PRECISION[]--coeffs from linear regression, out_table TEXT) RETURNS VARCHAR AS $$DECLARE insert_string VARCHAR
	Return insert string for robust linear regression.

CREATE OR REPLACE FUNCTION MADlib	robust_variance_linregr (source_table VARCHAR--name of input table, out_table VARCHAR--name of output table, dependent_variable VARCHAR--name of dependent variable, independent_variable VARCHAR--name of independent variable) RETURNS VOID AS $$BEGIN PERFORM MADlib.robust_variance_linregr(source_table
	Robust linear regression with default fit regression behaviour & no grouping.

Robust Linear Regression CREATE OR REPLACE FUNCTION MADlib	robust_variance_linregr (source_table VARCHAR--name of input table, out_table VARCHAR--name of output table, dependent_varname VARCHAR--name of dependent variable, input_independent_varname VARCHAR--name of independent variable, input_group_cols VARCHAR--grouping columns) RETURNS VOID AS $$DECLARE insert_string VARCHAR
	Robust linear regression function subcall.

aggregate robust_logregr_result	robust_logregr (boolean dependentVariable, float8[] independentVariables, float8[] coef)
	Compute robust regression diagnostic statistics for logistic regression. More...

CREATE FUNCTION MADlib	__internal_get_robust_logregr_result (source_table VARCHAR--name of input table, dependent_varname VARCHAR--name of dependent variable, independent_varname VARCHAR--name of independent variable, logregr_coeffs DOUBLE PRECISION[]--coeffs from logear regression) RETURNS MADlib.robust_logregr_result AS $$DECLARE robust_value MADlib.robust_logregr_result
	Return robust logistic regression estimates given a set of coefficients.

CREATE FUNCTION MADlib	__internal_get_robust_logregr_insert_string (robust_log_rst MADlib.robust_logregr_result, out_table TEXT) RETURNS VARCHAR AS $$DECLARE insert_string VARCHAR
	Return insert string for robust logistic regression.

void	robust_variance_logregr (varchar source_table, varchar out_table, varchar dependent_varname, varchar input_independent_varname, varchar input_group_cols, integer max_iter, varchar optimizer, float8 tolerance, boolean print_warnings)
	The robust logistic regression function. More...

CREATE OR REPLACE FUNCTION MADlib	robust_variance_logregr (source_table VARCHAR,--name of input table out_table VARCHAR,--name of output table dependent_variable VARCHAR,--name of dependent variable independent_variable VARCHAR,--name of independent variable input_group_cols VARCHAR--grouping columns) RETURNS VOID AS $$BEGIN PERFORM MADlib.robust_variance_logregr(source_table
	Robust logistic function subcall.

CREATE OR REPLACE FUNCTION MADlib	robust_variance_logregr (source_table VARCHAR--name of input table, out_table VARCHAR--name of output table, dependent_variable VARCHAR--name of dependent variable, independent_variable VARCHAR--name of independent variable) RETURNS VOID AS $$BEGIN PERFORM MADlib.robust_variance_logregr(source_table
	Robust logistic regression with default fit regression behavior, and no grouping,.

Detailed Description

Date: January 2011

See Also: Calculates robust statistics for various regression models.

Definition in file robust.sql_in.

Function Documentation

aggregate robust_linregr_result robust_linregr	(	float8	dependentVariable,
		float8[]	independentVariables,
		float8[]	coef
	)

Parameters

dependentVariable	Column containing the dependent variable
independentVariables	Column containing the array of independent variables
coef	Column containing the array of the OLS coefficients (as obtained by linregr)

: To include an intercept in the model, set one coordinate in the independentVariables array to 1.

Returns

A composite value:

std_err FLOAT8[] - Array of huber-white standard errors, $ \mathit{se}(c_1), \dots, \mathit{se}(c_k) $
t_stats FLOAT8[] - Array of t-statistics, $ \boldsymbol t $
p_values FLOAT8[] - Array of p-values, $ \boldsymbol p $

Usage:

Get all the diagnostic statistics:

 SELECT (robust_linregr(dependentVariable,
    independentVariables, coef)).*
    FROM (
    SELECT linregr(dependentVariable, independentVariables).coef
    ) AS ols_coef, sourceName as src;

Get a subset of the output columns, e.g., only the condition number and the array of p-values $ \boldsymbol p $:

SELECT (lr).robust_condition_no, (lr).robust_p_values
FROM (

 SELECT (robust_linregr(dependentVariable,
    independentVariables, coef)).*
    FROM (
    SELECT linregr(dependentVariable, independentVariables).coef
    ) AS ols_coef, sourceName as src
) AS subq;

Definition at line 320 of file robust.sql_in.

aggregate robust_logregr_result robust_logregr	(	boolean	dependentVariable,
		float8[]	independentVariables,
		float8[]	coef
	)

Parameters

dependentVariable	Column containing the dependent variable
independentVariables	Column containing the array of independent variables
coef	Column containing the array of the coefficients (as obtained by logregr)

: To include an intercept in the model, set one coordinate in the independentVariables array to 1.

Returns

A composite value:

coef FLOAT8[] - The coefficients for the regression
std_err FLOAT8[] - Array of huber-white standard errors, $ \mathit{se}(c_1), \dots, \mathit{se}(c_k) $
z_stats FLOAT8[] - Array of Wald z-statistics, $ \boldsymbol t $
p_values FLOAT8[] - Array of p-values, $ \boldsymbol p $

Usage:

Get all the diagnostic statistics:

 SELECT robust_logregr(dependentVariable,
 independentVariables, coef)
 FROM dataTable;

Definition at line 751 of file robust.sql_in.

void robust_variance_logregr	(	varchar	source_table,
		varchar	out_table,
		varchar	dependent_varname,
		varchar	input_independent_varname,
		varchar	input_group_cols,
		integer	max_iter,
		varchar	optimizer,
		float8	tolerance,
		boolean	print_warnings
	)

Parameters

source_table	String identifying the input table
out_table	String identifying the output table to be created
dependent_varname	Column containing the dependent variable
independent_varname	Column containing the array of independent variables
input_group_cols	Columns to group by.
max_iter	Integer identifying the maximum iterations used by the logistic regression solver. Default is 20.
optimizer	String identifying the optimizer used in the logistic regression. See the documentation in the logistic regression for the available options. Default is irls.
tolerance	Float identifying the tolerance of the logistic regression optimizer. Default is 0.0001.
print_warnings	Boolean specifying if the regression fit should print any warning messages. Default is false.

: To include an intercept in the model, set one coordinate in the independent_varname array to 1.

Returns

A composite value:

std_err FLOAT8[] - Array of huber-white standard errors, $ \mathit{se}(c_1), \dots, \mathit{se}(c_k) $
t_stats FLOAT8[] - Array of t-statistics, $ \boldsymbol t $
p_values FLOAT8[] - Array of p-values, $ \boldsymbol p $

Usage:

For function summary information. Run sql> select robust_variance_logregr('help'); OR sql> select robust_variance_logregr(); OR sql> select robust_variance_logregr('?'); For function usage information. Run sql> select robust_variance_logregr('usage');

Compute the coefficients, and the get the robust diagnostic statistics:

   select robust_variance_logregr(source_table, out_table, regression_type, dependentVariable, independentVariables, NULL );

If the coefficients are already known, they can be provided directly

select robust_variance_logregr(source_table, out_table, regression_type, dependentVariable, independentVariables, coef );

Definition at line 936 of file robust.sql_in.

Functions

Detailed Description

Function Documentation