Computes matrix that expands a single variable into the terms needed
 to fit a restricted cubic spline (natural spline) function using the
 truncated power basis. Two normalization options are given for
 somewhat reducing problems of ill-conditioning. The antiderivative
 function can be optionally created. If knot locations are not given,
 they will be estimated from the marginal distribution of <code>x</code>.

rcspline.eval

Contains many functions useful for data
analysis, high-level graphics, utility operations, functions for
computing sample size and power, simulation, importing and annotating datasets,
imputing missing values, advanced table making, variable clustering,
character string manipulation, conversion of R objects to Typst, LaTeX, and html code,
recoding variables, caching, simplified parallel computing, encrypting and decrypting data using a safe workflow, general moving window statistical estimation, and assistance in interpreting principal component analysis.

Frank Harrell Jr

Hmisc

Harrell Miscellaneous

Frank E Harrell Jr

Cole Beck

Charles Dupont

rcspline.eval function

Restricted Cubic Spline Design Matrix — rcspline.eval

<dl>

 <dt>x</dt>
<dd>a vector representing a predictor variable</dd>

 <dt>knots</dt>
<dd>knot locations. If not given, knots will be estimated using default
 quantiles of <code>x</code>. For 3 knots, the outer quantiles used are 0.10
 and 0.90. For 4-6 knots, the outer quantiles used are 0.05 and
 0.95. For \(\code{nk}&gt;6\), the outer quantiles are 0.025 and 0.975. The
 knots are equally spaced between these on the quantile scale. For
 fewer than 100 non-missing values of <code>x</code>, the outer knots are
 the 5th smallest and largest <code>x</code>.</dd>

 <dt>nk</dt>
<dd>number of knots. Default is 5. The minimum value is 3.</dd>

 <dt>inclx</dt>
<dd>set to <code>TRUE</code> to add <code>x</code> as the first column of the
 returned matrix</dd>

 <dt>knots.only</dt>
<dd>return the estimated knot locations but not the expanded matrix</dd>

 <dt>type</dt>
<dd><samp>"ordinary"</samp> to fit the function, <samp>"integral"</samp> to fit its
 anti-derivative.</dd>

 <dt>norm</dt>
<dd><samp>0</samp> to use the terms as originally given by <cite>Devlin and
 Weeks (1986)</cite>, <samp>1</samp> to normalize non-linear terms by the cube
 of the spacing between the last two knots, <samp>2</samp> to normalize by
 the square of the spacing between the first and last knots (the
 default). <code>norm=2</code> has the advantage of making all nonlinear
 terms beon the x-scale.</dd>

 <dt>rpm</dt>
<dd>If given, any <code>NA</code>s in <code>x</code> will be replaced with the value
 <code>rpm</code> after estimating any knot locations.</dd>

 <dt>pc</dt>
<dd>Set to <code>TRUE</code> to replace the design matrix with orthogonal
 (uncorrelated) principal components computed on the scaled, centered
 design matrix</dd>

	<dt>fractied</dt>
<dd>If the fraction of observations tied at the lowest and/or highest
		values of <code>x</code> is greater than or equal to <code>fractied</code>, the
		algorithm attempts to use a different algorithm for knot finding
		based on quantiles of <code>x</code> after excluding the one or two values
		with excessive ties. And if the number of unique <code>x</code> values
		excluding these values is small, the unique values will be used as
		the knots. If the number of knots to use other than these exterior
		values is only one, that knot will be at the median of the
		non-extreme <code>x</code>. This algorithm is not used if any interior
		values of <code>x</code> also have a proportion of ties equal to or
		exceeding <code>fractied</code>.</dd>

</dl>

rcspline.eval: Restricted Cubic Spline Design Matrix

Description

Usage

Arguments

Value

References

See Also

Examples