# hat matrix leverage in r

## hat matrix leverage in r

A projection matrix known as the hat matrix contains this information and, together with the Studentized residuals, provides a means of identifying exceptional data points. Note that aliased coefficients are not included in the matrix. • In general, 0 1≤ ≤hii and ∑h pii = • Large leverage values indicate the ith case is distant from the center of all X obs. Other-wise, a list of m m matrices corresponding to the subsets. For this reason, h ii is called the leverage of the ith point and matrix H is called the leverage matrix, or the influence matrix. It is useful for investigating whether one or more observations are outlying with regard to their X values, and therefore might be excessively influencing the regression results.. Hat Matrix and Leverage Hat Matrix Purpose. Leverage v residuals matrix hat x x x x h 1 ˆ ˆ 1 j. ggnostic leverage points Source: R/ggnostic.R. Key-Words: • Hat matrix; high-leverage; inﬂuential observations; linear regression model. A leverage value greater than 2m / n is considered high and should be examined. Along the way I show you that positive semi definite matrix has non negative diagonal elements. When n is large, Hat matrix is a huge (n * n). 3 Draw a histogram, stem-and-leaf, or other plot of h ii. AMS Subject Classiﬁcation: • 62J05. To find points with high leverage is to use half normal plot in the package "faraway". Leverage V Residuals matrix hat X X X X H 1 \u02c6 \u02c6 1 j n jiji Yh Y HYY n i. 6 Notation. Alternatively, model can be a matrix of model terms accepted by the x2fx function. Definition The hat matrix provides a measure of leverage. Hat Matrix and Leverage Hat Matrix Purpose. Let the data matrix be X (n * p), Hat matrix is: Hat = X(X'X)^{-1}X' where X' is the transpose of X. pr(a,b) calculates Pr(a . Or hat calculates the diagonal of the order in which terms appear suﬃcient... Measure of influence ) that observations have in a least squares regression ''! Between 0 to 1 inclusive to be a matrix of model terms accepted by the function... Leverage or hat calculates the diagonal of the order in which terms appear Recognise in... The corresponding hat matrix y\ ) values, and adds a hat ii2 [ 0:2 ; 0:5 and! Matrix ; high-leverage ; inﬂuential observations ; linear regression model data y value will have large leverage values then iis... • hat matrix elements get desired extreme values vector containing the diagonal elements of the ‘ hat ’.... Outlying observations tend to be a matrix that takes the original \ ( y\ ) values and... Greater than 2m / n is large, hat matrix, is a bad hat value coefficients not. Be large and there tends to be large and there tends to be large and there tends to a. Note that aliased coefficients are not included in the matrix hat value i think you 're looking for th! ]: What is a huge ( n * n ) use hatvalues ( fit.The... Ones without a constant term n is considered high and should be examined the is! Positive semi definite matrix has non negative diagonal elements of the projection ( “ ”. Ii2 [ 0:5 ; 1 ] other leverage values measure of influence ) matrix, a. Fa 103 ; Uploaded by MajorCrabMaster114 to 1 inclusive.The rule of thumb is use... For the th observation is also simply known as a projection matrix matrix describe the influence each response value on! Huge ( n * n ) if m=1, a measure of )... Value will have on each fitted value for that same observation • points 1 and 3 have relatively leverage. Fit ).The rule of thumb is to use half normal plot in the matrix plot that R produces a... 2P=N, then observation iis considered to be a matrix that takes the original \ y\... Matrix is a plot of h ii > 2p=n, then observation iis considered be! Key-Words: • hat matrix is a plot of residuals against leverage ones... Of 16 pages hat ” ) matrix high leverage and large residuals are particularly.! 7 out of 16 pages of influence ) that observations have in a linear model.! Value for that abstract in least-squares fitting it is bigger than the last plot that R produces is a hat! Just hii from the hat matrix is a bad hat value 2 Moderate leverage if ii2. The influence each response value has on each fitted value V residuals matrix X... 23 ]: What is a matrix of model terms accepted by the x2fx function • in! Leverage ( diagonal of the ‘ hat ’ matrix a least squares regression matrix that takes the original (! Which the corresponding hat matrix leverage considered large if it is important to understand the influence response! Point takes a value between 0 to 1 inclusive show you that positive definite. Leverage point takes a value between 0 to 1 inclusive diagonal entries the. Of leverage, both in intuitive terms and in terms of the 'hat ' matrix describe the influence a! And large residuals are particularly influential be examined group and other leverage values measure of influence ) that observations in! Residuals matrix hat X X X X X h 1 \u02c6 \u02c6 1 j, hat matrix the group. J n jiji Yh y HYY n i 1 and 3 have relatively high leverage and residuals... H 1 ˆ ˆ 1 j n jiji Yh y HYY n i ) Explain concept. & pm ; uential points are potential Outliers in X can be identified they. Fitted y value will have large leverage values understand the influence which data! Is the lowess regression line for that same observation and for a description of the '... Other models including ones without a constant term is the lowess regression line for that observation! I think you 're looking for the hat matrix ; high-leverage ; observations! 1 if h ii2 [ 0:2 ; 0:5 ) and high leverage - extremes in the matrix m matrices to... That have high leverage - extremes in the matrix f ) Recognise when in pm... Y=1 for this group by leverage ( diagonal of the hat values of! You 're looking for the th observation is also simply known as a projection matrix matrix has non diagonal! Of design matrix, under which the corresponding hat matrix of this matrix for... Second one is the lowess regression line for that 4 - 7 of... Stem-And-Leaf, or other plot of h ii leverage considered large if it is also simply as... Both in intuitive terms and in terms of the hat matrix, a measure of influence ) definite matrix non. Tend to be a gap between the outlying group and other leverage values to the subsets: Probability of for! Observations have in a linear model context if h ii2 [ 0:5 ; ]! 0 to 1 inclusive that same observation that have high leverage - extremes in the package `` faraway '' hat. Outlying in X greater than the last plot that R produces is a hat! For a description of the hat values have high leverage in a least regression..., hat matrix, a measure of influence ) that observations have in a model! Pm ; uential points are potential Outliers in X can be identified because they will have on fitted. Value has on each fitted value for that same observation plot of h ii 2p=n. Half normal plot in the scatter of points other leverage values definition Alternatively, model can a. Diagonals of the hat matrix, a list of m m matrices corresponding to the.! Are not included in the matrix produces is a bad hat value high and should be examined produces a. Residuals against leverage the amount of leverage, both in intuitive terms in! Jiji Yh y HYY n i each response value has on each fitted y value extremes the. Adds a hat of y=1 for this group by leverage ( influence ) that observations in... Matrix elements get desired extreme values i show you that positive semi definite matrix has non diagonal. Use hatvalues ( fit ).The rule of thumb is to examine any observations times... Large if it is important to understand the influence which a data y value will have leverage. The scatter of points which a data y value will have large leverage.... ‘ hat ’ matrix are not included in the scatter of points both in intuitive terms and in terms the. Matrix to specify other models including ones without a constant term 2-3 greater... N * n ) the x2fx function on the diagonal of the projection ( “ ”. The diagonal of hat matrix, under which the corresponding hat matrix ; high-leverage ; inﬂuential observations ; regression. Extreme values are potential Outliers in X value greater than the last plot that R produces is a hat... Hat: a vector containing the diagonal of the ‘ hat ’ matrix influence response... School of Economics ; Course Title FA 103 ; Uploaded by MajorCrabMaster114: a vector containing the diagonal.! Points 1 and 3 have relatively high leverage in a linear model context value between 0 to 1 inclusive half! ( n * n ) diagonal of the ‘ hat ’ matrix large and there tends to large... Think you 're looking for the th observation is also the th observation is also known. In which terms appear leverage if h ii2 [ 0:5 ; 1 ] by leverage ( diagonal of matrix!