statsample-optimization 2.0.5 → 2.1.0

data.tar.gz.sig

@@ -1,2 +1,2 @@
-I�S��W�{��M�J�4���G�iC��ĎX�g/��D�!|8��
�Z�������eZ��n���k6����-�����ģ��*�
h���p�D�W���^�B��Fo��
-eh`™�n̍�ᢚ�րX�I��L
w�yc^����n�ރ����Zyg�=�*�������,< q�Y���Q[�S���A�.w�_�ࡴH��+�^(�g����elJ&nAm3�̼�q$$B�2�vf���
v����P˦e��e�
+��0��,g�h��d��U���8^�ۡb�1-o�<YS
`��1�V�����:���Yz�/"!Y9�X�?�=���P��fxP�(;\u�]�
71�����T����@�����<N[�j�%X�<u��X
+	7

data/History.txt

@@ -1,8 +1,14 @@
+=== 2.1.0 / 2010-12-22
+* Added tetrachoric correlation support using fortran to c version of AS116
+* Added native compilation support
+
 === 2.0.3 / 2010-06-14
 * All extensions dependencies of statsample moved to statsample-optimization
+
 === 2.0.0 / 2009-08-21
  * Uses rcov internal configuration
-=== 1.0.0 / 2009-08-02
+
+ === 1.0.0 / 2009-08-02
 
  * Birthday!
 

data/Manifest.txt

@@ -2,6 +2,8 @@ History.txt
 Manifest.txt
 README.txt
 Rakefile
+ext/statsamplert/as116.c
+ext/statsamplert/as116.h
 ext/statsamplert/extconf.rb
 ext/statsamplert/statsamplert.c
 test/test_statsample_optimization.rb

data/Rakefile

@@ -8,11 +8,11 @@ Hoe.plugin :git
 
 task "test" => ["lib/statsamplert.so"]
 spec=Hoe.spec 'statsample-optimization' do
-  self.version="2.0.5"
+  self.version="2.1.0"
   self.spec_extras[:extensions] = ["ext/statsamplert/extconf.rb"]
   self.rubyforge_name = 'ruby-statsample'
   self.developer('Claudio Bustos', 'clbustos_at_gmail.com')
-  self.extra_deps << ["statsample","~>0.15.1"] << ["statistics2", "~>0.54"] << ["gsl", "~>1.12.109"]
+  self.extra_deps << ["statsample","~>0.17"] << ["statistics2", "~>0.54"] << ["gsl", "~>1.14"]
   self.extra_deps  << ["rake-compiler"]
 
 end

data/ext/statsamplert/as116.c

@@ -0,0 +1,638 @@
+/* as116.f -- translated by f2c (version 20090411).
+   You must link the resulting object file with libf2c:
+	on Microsoft Windows system, link with libf2c.lib;
+	on Linux or Unix systems, link with .../path/to/libf2c.a -lm
+	or, if you install libf2c.a in a standard place, with -lf2c -lm
+	-- in that order, at the end of the command line, as in
+		cc *.o -lf2c -lm
+	Source for libf2c is in /netlib/f2c/libf2c.zip, e.g.,
+
+		http://www.netlib.org/f2c/libf2c.zip
+*/
+#include "math.h"
+#include "as116.h"
+
+typedef int logical;
+
+
+#define dabs(x) (double)abs(x)
+#define abs(x) ((x) >= 0 ? (x) : -(x))
+
+#define TRUE_ (1)
+#define FALSE_ (0)
+
+
+static double alnorm_(double *x, logical *upper);
+static double ppnd_(double *p, int *ier);
+
+
+
+
+
+
+/* Table of constant values */
+
+static logical c_false = FALSE_;
+
+/* This file includes the Applied Statistics algorithm AS 66 for calculating */
+/* the tail area under the normal curve, and two alternative routines which */
+/* give higher accuracy.   The latter have been contributed by Alan Miller of */
+/* CSIRO Division of Mathematics & Statistics, Clayton, Victoria.   Notice */
+/* that each function or routine has different call arguments. */
+
+
+static double alnorm_(double *x, logical *upper) {
+    /* Initialized data */
+
+    static double zero = 0.;
+    static double a1 = 5.75885480458;
+    static double a2 = 2.62433121679;
+    static double a3 = 5.92885724438;
+    static double b1 = -29.8213557807;
+    static double b2 = 48.6959930692;
+    static double c1 = -3.8052e-8;
+    static double c2 = 3.98064794e-4;
+    static double c3 = -.151679116635;
+    static double c4 = 4.8385912808;
+    static double c5 = .742380924027;
+    static double one = 1.;
+    static double c6 = 3.99019417011;
+    static double d1 = 1.00000615302;
+    static double d2 = 1.98615381364;
+    static double d3 = 5.29330324926;
+    static double d4 = -15.1508972451;
+    static double d5 = 30.789933034;
+    static double half = .5;
+    static double ltone = 7.;
+    static double utzero = 18.66;
+    static double con = 1.28;
+    static double p = .398942280444;
+    static double q = .39990348504;
+    static double r__ = .398942280385;
+
+    /* System generated locals */
+    double ret_val;
+
+
+    /* Local variables */
+    static double y, z__;
+    static logical up;
+
+
+/*         Algorithm AS66 Applied Statistics (1973) vol22 no.3 */
+
+/*       Evaluates the tail area of the standardised normal curve */
+/*       from x to infinity if upper is .true. or */
+/*       from minus infinity to x if upper is .false. */
+
+/* *** machine dependent constants */
+
+    up = *upper;
+    z__ = *x;
+    if (z__ >= zero) {
+	goto L10;
+    }
+    up = ! up;
+    z__ = -z__;
+L10:
+    if (z__ <= ltone || (up && z__ <= utzero)) {
+	goto L20;
+    }
+    ret_val = zero;
+    goto L40;
+L20:
+    y = half * z__ * z__;
+    if (z__ > con) {
+	goto L30;
+    }
+
+    ret_val = half - z__ * (p - q * y / (y + a1 + b1 / (y + a2 + b2 / (y + a3)
+	    )));
+    goto L40;
+L30:
+    ret_val = r__ * exp(-y) / (z__ + c1 + d1 / (z__ + c2 + d2 / (z__ + c3 + 
+	    d3 / (z__ + c4 + d4 / (z__ + c5 + d5 / (z__ + c6))))));
+L40:
+    if (! up) {
+	ret_val = one - ret_val;
+    }
+    return ret_val;
+} /* alnorm_ */
+
+static double ppnd_(double *p, int *ier)
+{
+    /* Initialized data */
+
+    static double split = .42;
+    static double a0 = 2.50662823884;
+    static double a1 = -18.61500062529;
+    static double a2 = 41.39119773534;
+    static double a3 = -25.44106049637;
+    static double b1 = -8.4735109309;
+    static double b2 = 23.08336743743;
+    static double b3 = -21.06224101826;
+    static double b4 = 3.13082909833;
+    static double c0 = -2.78718931138;
+    static double c1 = -2.29796479134;
+    static double c2 = 4.85014127135;
+    static double c3 = 2.32121276858;
+    static double d1 = 3.54388924762;
+    static double d2 = 1.63706781897;
+    static double zero = 0.;
+    static double one = 1.;
+    static double half = .5;
+
+    /* System generated locals */
+    double ret_val;
+
+
+    /* Local variables */
+    static double q, r__;
+
+
+/*       ALGORITHM AS 111, APPL.STATIST., VOL.26, 118-121, 1977. */
+
+/*       PRODUCES NORMAL DEVIATE CORRESPONDING TO LOWER TAIL AREA = P. */
+
+/* 	See also AS 241 which contains alternative routines accurate to */
+/* 	about 7 and 16 decimal digits. */
+
+
+    *ier = 0;
+    q = *p - half;
+    if (abs(q) > split) {
+	goto L10;
+    }
+
+/*       0.08 < P < 0.92 */
+
+    r__ = q * q;
+    ret_val = q * (((a3 * r__ + a2) * r__ + a1) * r__ + a0) / ((((b4 * r__ + 
+	    b3) * r__ + b2) * r__ + b1) * r__ + one);
+    return ret_val;
+
+/*       P < 0.08 OR P > 0.92, SET R = MIN(P,1-P) */
+
+L10:
+    r__ = *p;
+    if (q > zero) {
+	r__ = one - *p;
+    }
+    if (r__ <= zero) {
+	goto L20;
+    }
+    r__ = sqrt(-log(r__));
+    ret_val = (((c3 * r__ + c2) * r__ + c1) * r__ + c0) / ((d2 * r__ + d1) * 
+	    r__ + one);
+    if (q < zero) {
+	ret_val = -ret_val;
+    }
+    return ret_val;
+L20:
+    *ier = 1;
+    ret_val = zero;
+    return ret_val;
+} /* ppnd_ */
+
+/* Subroutine */ 
+
+int tetra(double *a, double *b, double *c__, double *d__, double *r__,
+	 double *sdr, double *sdzero, double *threshold_x, double *threshold_y, int *itype, int *ifault)
+{
+    /* Initialized data */
+
+    static double x[16] = { .9972638618f,.9856115115f,.9647622556f,.9349060759f,
+	    .8963211558f,.8493676137f,.794483796f,.7321821187f,.6630442669f,
+	    .5877157572f,.5068999089f,.4213512761f,.3318686023f,.2392873623f,
+	    .1444719616f,.0483076657f };
+    static double w[16] = { .00701861f,.0162743947f,.0253920653f,.0342738629f,
+	    .042835898f,.0509980593f,.0586840935f,.0658222228f,.0723457941f,
+	    .0781938958f,.0833119242f,.087652093f,.0911738787f,.0938443991f,
+	    .0956387201f,.0965400885f };
+    static double zero = 0.f;
+    static double one = 1.f;
+    static double two = 2.f;
+    static double four = 4.f;
+    static double six = 6.f;
+    static double half = .5f;
+    static double twopi = 6.28318531f;
+    static double sqt2pi = 2.50662827f;
+    static double rlimit = .9999f;
+    static double rcut = .95f;
+    static double uplim = 5.f;
+    static double const__ = 1e-36f;
+    static double chalf = 1e-18f;
+    static double conv = 1e-8f;
+    static double citer = 1e-6f;
+    static int niter = 25;
+
+    /* System generated locals */
+    double r__1, r__2, r__3;
+
+
+    /* Local variables */
+    static double aa, bb, cc, dd;
+    static int ie;
+    static double dr, va, vb, wa, wb, rr, sr, ss, pab, pac, cof, vaa, waa, pdf, 
+	    zab, zac, xla, xlb, sum, tot, amid;
+    static double prob;
+    static int iter;
+    static double xlen, term, rrsq, delta;
+    static int iquad;
+    static double tempa, tempb, deriv;
+    static int ksign, iterm;
+    static double rrest;
+    static int kdelta;
+    static double probaa, probac, probab;
+    static double rrprev, sumprv;
+
+
+/*        ALGORITHM  AS 116 APPL. STATIST. (1977) VOL.26, NO.3 */
+
+/*        TO COMPUTE THE TETRACHORIC CORRELATION (R) AND ITS STANDARD */
+/*        ERRORS (SDR AND SDZERO) FROM THE FREQUENCIES OF A 2*2 TABLE */
+/*        (A, B, C AND D) */
+/*        X AND W ARE CONSTANTS USED IN GAUSSIAN QUADRATURE */
+
+
+/*        INITIALIZATION */
+
+    *r__ = zero;
+    *sdzero = zero;
+    *sdr = zero;
+    *itype = 0;
+    *ifault = 0;
+    *threshold_x=zero;
+    *threshold_y=zero;
+    
+/*       CHECK IF ANY CELL FREQUENCY IS NEGATIVE */
+
+    if (*a < zero || *b < zero || *c__ < zero || *d__ < zero) {
+	goto L92;
+    }
+
+/*       CHECK IF ANY FREQUENCY IS ZERO AND SET KDELTA */
+
+    kdelta = 1;
+    delta = zero;
+    if (*a == zero || *d__ == zero) {
+	kdelta = 2;
+    }
+    if (*b == zero || *c__ == zero) {
+	kdelta += 2;
+    }
+
+/*        KDELTA=4 MEANS TABLE HAS ZERO ROW OR COLUMN, RUN IS TERMINATED */
+
+    switch (kdelta) {
+	case 1:  goto L4;
+	case 2:  goto L1;
+	case 3:  goto L2;
+	case 4:  goto L92;
+    }
+
+/*        DELTA IS 0.0, 0.5 OR -0.5 ACCORDING TO WHICH CELL IS ZERO */
+
+L1:
+    delta = half;
+    if (*a == zero && *d__ == zero) {
+	*r__ = -one;
+    }
+    goto L4;
+L2:
+    delta = -half;
+    if (*b == zero && *c__ == zero) {
+	*r__ = one;
+    }
+L4:
+    if (*r__ != zero) {
+	*itype = 3;
+    }
+
+/*        STORE FREQUENCIES IN  AA, BB, CC AND DD */
+
+    aa = *a + delta;
+    bb = *b - delta;
+    cc = *c__ - delta;
+    dd = *d__ + delta;
+    tot = aa + bb + cc + dd;
+
+/*        CHECK IF CORRELATION IS NEGATIVE, ZERO, POSITIVE */
+
+    if ((r__1 = aa * dd - bb * cc) < 0.f) {
+	goto L7;
+    } else if (r__1 == 0) {
+	goto L5;
+    } else {
+	goto L6;
+    }
+L5:
+    *itype = 4;
+
+/*        COMPUTE PROBABILITIES OF QUADRANT AND OF MARGINALS */
+/*        PROBAA AND PROBAC CHOSEN SO THAT CORRELATION IS POSITIVE. */
+/*        KSIGN INDICATES WHETHER QUADRANTS HAVE BEEN SWITCHED */
+
+L6:
+    probaa = aa / tot;
+    probac = (aa + cc) / tot;
+    ksign = 1;
+    goto L8;
+L7:
+    probaa = bb / tot;
+    probac = (bb + dd) / tot;
+    ksign = 2;
+L8:
+    probab = (aa + bb) / tot;
+
+/*        COMPUTE NORMAL DEVIATES FOR THE MARGINAL FREQUENCIES */
+/*        SINCE NO MARGINAL CAN BE ZERO, IE IS NOT CHECKED */
+
+    zac = ppnd_((double*)&probac, &ie);
+    zab = ppnd_((double*)&probab, &ie);
+
+    *threshold_x=zab;    
+    *threshold_y=zac;
+/* Computing 2nd power */
+    r__1 = zac;
+/* Computing 2nd power */
+    r__2 = zab;
+    ss = exp(-half * (r__1 * r__1 + r__2 * r__2)) / twopi;
+
+/*        WHEN R IS 0.0, 1.0 OR -1.0, TRANSFER TO COMPUTE SDZERO */
+
+    if (*r__ != zero || *itype > 0) {
+	goto L85;
+    }
+
+/*        WHEN MARGINALS ARE EQUAL, COSINE EVALUATION IS USED */
+
+    if (*a == *d__ && *b == *c__) {
+	goto L60;
+    }
+
+/*        INITIAL ESTIMATE OF CORRELATION IS YULES Y */
+
+/* Computing 2nd power */
+    r__2 = sqrt(aa * dd) - sqrt(bb * cc);
+    rr = r__2 * r__2 / (r__1 = aa * dd - bb * cc, dabs(r__1));
+    iter = 0;
+
+/*        IF RR EXCEEDS RCUT, GAUSSIAN QUADRATURE IS USED */
+
+L10:
+    if (rr > rcut) {
+	goto L40;
+    }
+
+/*        TETRACHORIC SERIES IS COMPUTED */
+
+/*        INITIALIZATION */
+
+    va = one;
+    vb = zac;
+    wa = one;
+    wb = zab;
+    term = one;
+    iterm = 0;
+    sum = probab * probac;
+    deriv = zero;
+    sr = ss;
+L15:
+    if (dabs(sr) > const__) {
+	goto L20;
+    }
+
+/*        RESCALE TERMS TO AVOID OVERFLOWS AND UNDERFLOWS */
+
+    sr /= const__;
+    va *= chalf;
+    vb *= chalf;
+    wa *= chalf;
+    wb *= chalf;
+
+/*        FORM SUM AND DERIVATIVE OF SERIES */
+
+L20:
+    dr = sr * va * wa;
+    sr = sr * rr / term;
+    cof = sr * va * wa;
+
+/*        ITERM COUNTS NO. OF CONSECUTIVE TERMS .LT. CONV */
+
+    ++iterm;
+    if (dabs(cof) > conv) {
+	iterm = 0;
+    }
+    sum += cof;
+    deriv += dr;
+    vaa = va;
+    waa = wa;
+    va = vb;
+    wa = wb;
+    vb = zac * va - term * vaa;
+    wb = zab * wa - term * waa;
+    term += one;
+    if (iterm < 2 || term < six) {
+	goto L15;
+    }
+
+/*        CHECK IF ITERATION CONVERGED */
+
+    if ((r__1 = sum - probaa, dabs(r__1)) > citer) {
+	goto L25;
+    }
+
+/*        ITERATION HAS CONVERGED, SET ITYPE */
+
+    *itype = term;
+    goto L70;
+
+/*        CALCULATE NEXT ESTIMATE OF CORRELATION */
+
+L25:
+    ++iter;
+
+/*        IF TOO MANY ITERATlONS, RUN IS TERMINATED */
+
+    if (iter >= niter) {
+	goto L93;
+    }
+    delta = (sum - probaa) / deriv;
+    rrprev = rr;
+    rr -= delta;
+    if (iter == 1) {
+	rr += half * delta;
+    }
+    if (rr > rlimit) {
+	rr = rlimit;
+    }
+    if (rr < zero) {
+	rr = zero;
+    }
+    goto L10;
+
+/*       GAUSSIAN QUADRATURE */
+
+L40:
+    if (iter > 0) {
+	goto L41;
+    }
+
+/*       INITIALIZATION, IF THIS IS FIRST ITERATION */
+
+    sum = probab * probac;
+    rrprev = zero;
+
+/*        INITIALIZATION */
+
+L41:
+    sumprv = probab - sum;
+    prob = bb / tot;
+    if (ksign == 2) {
+	prob = aa / tot;
+    }
+    *itype = 1;
+
+/*        LOOP TO FIND ESTIMATE OF CORRELATION */
+/*        COMPUTATION OF INTEGRAL (SUM) BY QUADRATURE */
+
+L42:
+/* Computing 2nd power */
+    r__1 = rr;
+    rrsq = sqrt(one - r__1 * r__1);
+    amid = half * (uplim + zac);
+    xlen = uplim - amid;
+    sum = zero;
+    for (iquad = 1; iquad <= 16; ++iquad) {
+	xla = amid + x[iquad - 1] * xlen;
+	xlb = amid - x[iquad - 1] * xlen;
+
+/*       TO AVOID UNDERFLOWS, TEMPA AND TEMPB ARE USED */
+
+	tempa = (zab - rr * xla) / rrsq;
+	if (tempa >= -six) {
+/* Computing 2nd power */
+	    r__1 = xla;
+	    sum += w[iquad - 1] * exp(-half * (r__1 * r__1)) * alnorm_((
+		    double*)&tempa, &c_false);
+	}
+	tempb = (zab - rr * xlb) / rrsq;
+	if (tempb >= -six) {
+/* Computing 2nd power */
+	    r__1 = xlb;
+	    sum += w[iquad - 1] * exp(-half * (r__1 * r__1)) * alnorm_((
+		    double*)&tempb, &c_false);
+	}
+/* L44: */
+    }
+    sum = sum * xlen / sqt2pi;
+
+/*        CHECK IF ITERATION HAS CONVERGED */
+
+    if ((r__1 = prob - sum, dabs(r__1)) <= citer) {
+	goto L70;
+    }
+    ++iter;
+
+/*         IF TOO MANY ITERATIONS, RUN IS TERMINATED */
+
+    if (iter >= niter) {
+	goto L93;
+    }
+
+/*         ESTIMATE CORRELATION FOR NEXT ITERATION BY LINEAR INTERPOLATION */
+
+    rrest = ((prob - sum) * rrprev - (prob - sumprv) * rr) / (sumprv - sum);
+
+/*        IS ESTIMATE POSITIVE AND LESS THAN UPPER LIMIT */
+
+    if (rrest > rlimit) {
+	rrest = rlimit;
+    }
+    if (rrest < zero) {
+	rrest = zero;
+    }
+    rrprev = rr;
+    rr = rrest;
+    sumprv = sum;
+
+/*        IF ESTIMATE HAS SAME VALUE ON TWO ITERATIONS, STOP ITERATION */
+
+    if (rr == rrprev) {
+	goto L70;
+    }
+    goto L42;
+
+/*        WHEN ALL MARGINALS ARE EQUAL THE COSINE FUNCTION IS USED */
+
+L60:
+    rr = -cos(twopi * probaa);
+    *itype = 2;
+
+/*        COMPUTE SDR */
+
+L70:
+    *r__ = rr;
+/* Computing 2nd power */
+    r__1 = *r__;
+    rrsq = sqrt(one - r__1 * r__1);
+    if (kdelta > 1) {
+	*itype = -(*itype);
+    }
+    if (ksign == 1) {
+	goto L71;
+    }
+    *r__ = -(*r__);
+    zac = -zac;
+    *threshold_y=zac;    
+    
+L71:
+/* Computing 2nd power */
+    r__1 = zac;
+/* Computing 2nd power */
+    r__2 = zab;
+/* Computing 2nd power */
+    r__3 = rrsq;
+    pdf = exp(-half * (r__1 * r__1 - two * *r__ * zac * zab + r__2 * r__2) / (
+	    r__3 * r__3)) / (twopi * rrsq);
+    r__1 = (zac - *r__ * zab) / rrsq;
+    pac = alnorm_((double*)&r__1, &c_false) - half;
+    r__1 = (zab - *r__ * zac) / rrsq;
+    pab = alnorm_((double*)&r__1, &c_false) - half;
+/* Computing 2nd power */
+    r__1 = pab;
+/* Computing 2nd power */
+    r__2 = pac;
+    *sdr = (aa + dd) * (bb + cc) / four + r__1 * r__1 * (aa + cc) * (bb + dd) 
+	    + r__2 * r__2 * (aa + bb) * (cc + dd) + two * pab * pac * (aa * 
+	    dd - bb * cc) - pab * (aa * bb - cc * dd) - pac * (aa * cc - bb * 
+	    dd);
+    if (*sdr < zero) {
+	*sdr = zero;
+    }
+    *sdr = sqrt(*sdr) / (tot * pdf * sqrt(tot));
+
+/*        COMPUTE SDZERO */
+
+L85:
+/* Computing 2nd power */
+    r__1 = tot;
+    *sdzero = sqrt((aa + bb) * (aa + cc) * (bb + dd) * (cc + dd) / tot) / (
+	    r__1 * r__1 * ss);
+    if (*r__ == zero) {
+	*sdr = *sdzero;
+    }
+    goto L99;
+
+/*        ERROR TERMINATIONS */
+
+L92:
+    *ifault = 1;
+L93:
+    ++(*ifault);
+
+L99:
+    return 0;
+} /* tetra_ */
+

data/ext/statsamplert/as116.h

@@ -0,0 +1,18 @@
+#ifndef AS116_H
+#define AS116_H
+
+// A, B,C__, D__ -> the four cell frequencies [[A,B],[C,D]]
+// R__ -> The tetrachoric correlation
+// sdr -> The standard deviation of r
+// sdzero -> standard deviation of r, appropiate to test that correlation is 0
+// itype-> numerical method used
+// ifault-> error indicator 
+//  0->normal
+//  1->iteration did not converge
+//  2->tabla has at most one nonzero row or column
+
+int tetra(double *a, double *b, double *c__, double *d__, double *r__,
+	 double *sdr, double *sdzero, double *tx, double *ty, int *itype, int *ifault);
+
+#endif
+

data/ext/statsamplert/statsamplert.c

@@ -1,12 +1,16 @@
 #include <ruby.h>
+#include "as116.h"
 /**
 * :stopdoc:
 */
+
 void Init_statsamplert();
 VALUE statsample_frequencies(VALUE self, VALUE data);
 VALUE statsample_set_valid_data_intern(VALUE self, VALUE vector);
 VALUE statsample_case_as_hash(VALUE self, VALUE ds, VALUE index);
 VALUE statsample_case_as_array(VALUE self, VALUE ds, VALUE index);
+VALUE statsample_tetrachoric(VALUE self, VALUE a, VALUE b, VALUE c, VALUE d);
+
 void Init_statsamplert()
 {
 	VALUE mStatsample;
@@ -32,6 +36,7 @@ void Init_statsamplert()
     rb_define_module_function(mSTATSAMPLE__,"set_valid_data_intern", statsample_set_valid_data_intern, 1);
     rb_define_module_function(mSTATSAMPLE__,"case_as_hash",statsample_case_as_hash,2);
     rb_define_module_function(mSTATSAMPLE__,"case_as_array",statsample_case_as_array,2);
+    rb_define_module_function(mSTATSAMPLE__,"tetrachoric",statsample_tetrachoric,4);
 
 }
 
@@ -78,7 +83,7 @@ VALUE statsample_set_valid_data_intern(VALUE self, VALUE vector) {
 VALUE statsample_frequencies(VALUE self, VALUE data) {
     VALUE h;
     VALUE val;
-     long len;
+    long len;
     long i;
 
 	Check_Type(data,T_ARRAY);
@@ -128,3 +133,38 @@ VALUE statsample_case_as_array(VALUE self, VALUE ds, VALUE index) {
     return ar;
 }
 
+
+
+VALUE statsample_tetrachoric(VALUE self, VALUE a, VALUE b, VALUE c, VALUE d) {
+    VALUE h=rb_hash_new();
+    double pa;
+    double pb;
+    double pc;
+    double pd;
+    double r;
+	double sdr; 
+    double sdzero;
+	double t_x; 
+    double t_y;
+    
+    int itype;
+    int ifault;
+    int result;
+    pa=NUM2DBL(a);
+    pb=NUM2DBL(b);
+    pc=NUM2DBL(c);
+    pd=NUM2DBL(d);
+    
+    result= tetra(&pa,&pb, &pc, &pd, &r,
+	 &sdr, &sdzero, &t_x,&t_y, &itype, &ifault);
+    rb_hash_aset(h, rb_str_new_cstr("r"), DBL2NUM(r));
+    rb_hash_aset(h, rb_str_new_cstr("sdr"), DBL2NUM(sdr));
+    rb_hash_aset(h, rb_str_new_cstr("sdzero"), DBL2NUM(sdzero));
+    rb_hash_aset(h, rb_str_new_cstr("threshold_x"), DBL2NUM(t_x));
+    rb_hash_aset(h, rb_str_new_cstr("threshold_y"), DBL2NUM(t_y));
+    rb_hash_aset(h, rb_str_new_cstr("itype"), INT2NUM(itype));
+    rb_hash_aset(h, rb_str_new_cstr("ifault"), INT2NUM(ifault));
+    
+    return h;
+    
+}

data/test/test_statsample_optimization.rb

@@ -39,5 +39,15 @@ class TestRubyStatsampleOpt < Test::Unit::TestCase
     ds={"a"=>a,"b"=>b}.to_dataset
     assert_equal({"a"=>1,"b"=>4},Statsample::STATSAMPLE__.case_as_hash(ds,0))
   end
+  def test_tetrachoric
+    assert(Statsample::STATSAMPLE__.respond_to?(:tetrachoric))
+    a,b,c,d = 30,40,70,20
+    result=Statsample::STATSAMPLE__.tetrachoric(a,b,c,d)
+    assert_in_delta(-0.53980, result['r'], 0.0001)
+    assert_in_delta(0.09940, result['sdr'], 0.0001)
+    assert_in_delta(-0.15731, result['threshold_x'], 0.0001)
+    assert_in_delta(0.31864, result['threshold_y'], 0.0001)
+    
+  end
 
 end

metadata

@@ -1,13 +1,13 @@
 --- !ruby/object:Gem::Specification 
 name: statsample-optimization
 version: !ruby/object:Gem::Version 
-  hash: 5
+  hash: 11
   prerelease: false
   segments: 
   - 2
+  - 1
   - 0
-  - 5
-  version: 2.0.5
+  version: 2.1.0
 platform: ruby
 authors: 
 - Claudio Bustos
@@ -36,7 +36,7 @@ cert_chain:
   rpP0jjs0
   -----END CERTIFICATE-----
 
-date: 2010-11-13 00:00:00 -03:00
+date: 2010-12-22 00:00:00 -03:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency 
@@ -47,12 +47,11 @@ dependencies:
     requirements: 
     - - ~>
       - !ruby/object:Gem::Version 
-        hash: 33
+        hash: 41
         segments: 
         - 0
-        - 15
-        - 1
-        version: 0.15.1
+        - 17
+        version: "0.17"
   type: :runtime
   version_requirements: *id001
 - !ruby/object:Gem::Dependency 
@@ -78,12 +77,11 @@ dependencies:
     requirements: 
     - - ~>
       - !ruby/object:Gem::Version 
-        hash: 253
+        hash: 19
         segments: 
         - 1
-        - 12
-        - 109
-        version: 1.12.109
+        - 14
+        version: "1.14"
   type: :runtime
   version_requirements: *id003
 - !ruby/object:Gem::Dependency 
@@ -101,37 +99,21 @@ dependencies:
   type: :runtime
   version_requirements: *id004
 - !ruby/object:Gem::Dependency 
-  name: rubyforge
+  name: hoe
   prerelease: false
   requirement: &id005 !ruby/object:Gem::Requirement 
     none: false
     requirements: 
     - - ">="
       - !ruby/object:Gem::Version 
-        hash: 7
+        hash: 47
         segments: 
         - 2
+        - 8
         - 0
-        - 4
-        version: 2.0.4
+        version: 2.8.0
   type: :development
   version_requirements: *id005
-- !ruby/object:Gem::Dependency 
-  name: hoe
-  prerelease: false
-  requirement: &id006 !ruby/object:Gem::Requirement 
-    none: false
-    requirements: 
-    - - ">="
-      - !ruby/object:Gem::Version 
-        hash: 19
-        segments: 
-        - 2
-        - 6
-        - 2
-        version: 2.6.2
-  type: :development
-  version_requirements: *id006
 description: |-
   Install gsl, statistics2 and provides a C extension to optimize the following methods
   
@@ -154,6 +136,8 @@ files:
 - Manifest.txt
 - README.txt
 - Rakefile
+- ext/statsamplert/as116.c
+- ext/statsamplert/as116.h
 - ext/statsamplert/extconf.rb
 - ext/statsamplert/statsamplert.c
 - test/test_statsample_optimization.rb