<< readxlsx Basic functions repmat >>
Grocer >> Basic functions > remNaNs_spline

remNaNs_spline

fills %nan values in an matrix

CALLING SEQUENCE

[X,indNaN]=remNaNs_spline(X,method,k)

PARAMETERS

Input

* X = a (T x k) matrix of values, data in columns

* method = a scalar, the way Na are filled:

   . 1: replace all the missing values with a digital filter

   . 2: Replaces missing values after removing trailing and leading NA rows (a row is 'missing' if >80% is NaN)

   . 3: only remove rows with leading and closing zeros

   . 4: replace missing values after removing trailing and leading rows with all values %nan

* k = an integer, that controls the rational transfer function argument's numerator (the denominator is set to 1). More specifically, the numerator of the filter takes the form: ones(2*k+1,1)./[1;1/z;...;1/z^k]/(2*k+1)

Output

* X = the (T x k) matrix of values, with NA values repplaced

* indNaN = a (T x k) matrix, A matrix indicating the location for missing values

DESCRIPTION

Fills missing values (%nan) in an matrix.

EXAMPLE

global GROCERDIR ;
load(GROCERDIR+'\data\xnan.dat')

// replace NA values in X with a digital filter of order 5 (=2*2+1) 
// in the numerator
X_new1=remNaNs_spline(X,1,2)

// remove NA values in the last 2 rows and replace other NA values in X with 
// a digital filter of order 5 (=2*2+1) in the numerator applied to
// interpolated data
X_new2=remNaNs_spline(X,2,2)

// only remove NA values in the last 2 rows 
X_new3=remNaNs_spline(X,3)

// replace other NA values in X with 
// a digital filter of order 5 (=2*2+1) in the numerator applied to
// interpolated data
X_new4=remNaNs_spline(X,4,2)

AUTHOR

Éric Dubois 2026

Report an issue
<< readxlsx Basic functions repmat >>