@math.gl/geospatial 5.0.0-alpha.8 → 5.0.0-beta.1

package/src/vendor/geographiclib-geodesic/runtime.js

@@ -0,0 +1,2339 @@
+// math.gl
+// SPDX-License-Identifier: MIT
+// SPDX-FileCopyrightText: Copyright (c) 2011-2022 Charles Karney
+// SPDX-FileCopyrightText: Copyright (c) vis.gl contributors
+// SPDX-FileComment: Original math.gl follow-up evaluates short-edge cosine-series differences without cancellation; qualified by independent Decimal small-area controls. GeographicLib geodesic Math.js, Geodesic.js and GeodesicLine.js, pinned at 9c4036d6242aa997e08e43a20df789e2007fbe6f. Combined into an ESM module; version template fields set to 2.2.0. See ../../../GEOGRAPHICLIB-LICENSE.txt.
+
+/*
+ * Math.js
+ * Transcription of Math.hpp, Constants.hpp, and Accumulator.hpp into
+ * JavaScript.
+ *
+ * Copyright (c) Charles Karney (2011-2021) <karney@alum.mit.edu> and licensed
+ * under the MIT/X11 License.  For more information, see
+ * https://geographiclib.sourceforge.io/
+ */
+
+/**
+ * @namespace geodesic
+ * @description The parent namespace for the following modules:
+ * - {@link module:geodesic/Geodesic geodesic/Geodesic} The main
+ *   engine for solving geodesic problems via the
+ *   {@link module:geodesic/Geodesic.Geodesic Geodesic} class.
+ * - {@link module:geodesic/GeodesicLine geodesic/GeodesicLine}
+ *   computes points along a single geodesic line via the
+ *   {@link module:geodesic/GeodesicLine.GeodesicLine GeodesicLine}
+ *   class.
+ * - {@link module:geodesic/PolygonArea geodesic/PolygonArea}
+ *   computes the area of a geodesic polygon via the
+ *   {@link module:geodesic/PolygonArea.PolygonArea PolygonArea}
+ *   class.
+ * - {@link module:geodesic/Constants geodesic/Constants} defines
+ *   constants specifying the version numbers and the parameters for the WGS84
+ *   ellipsoid.
+ *
+ * The following modules are used internally by the package:
+ * - {@link module:geodesic/Math geodesic/Math} defines various
+ *   mathematical functions.
+ * - {@link module:geodesic/Accumulator geodesic/Accumulator}
+ *   interally used by
+ *   {@link module:geodesic/PolygonArea.PolygonArea PolygonArea} (via the
+ *   {@link module:geodesic/Accumulator.Accumulator Accumulator} class)
+ *   for summing the contributions to the area of a polygon.
+ */
+
+// To allow swap via [y, x] = [x, y]
+/* jshint esversion: 6 */
+
+var geodesic = {};
+geodesic.Constants = {};
+geodesic.Math = {};
+geodesic.Accumulator = {};
+
+(function(
+  /**
+   * @exports geodesic/Constants
+   * @description Define constants defining the version and WGS84 parameters.
+   */
+  c) {
+  "use strict";
+
+  /**
+   * @constant
+   * @summary WGS84 parameters.
+   * @property {number} a the equatorial radius (meters).
+   * @property {number} f the flattening.
+   */
+  c.WGS84 = { a: 6378137, f: 1/298.257223563 };
+  /**
+   * @constant
+   * @summary an array of version numbers.
+   * @property {number} major the major version number.
+   * @property {number} minor the minor version number.
+   * @property {number} patch the patch number.
+   */
+  c.version = { major: 2,
+                minor: 2,
+                patch: 0 };
+  /**
+   * @constant
+   * @summary version string
+   */
+  c.version_string = "2.2.0";
+})(geodesic.Constants);
+
+(function(
+  /**
+   * @exports geodesic/Math
+   * @description Some useful mathematical constants and functions (mainly for
+   *   internal use).
+   */
+  m) {
+  "use strict";
+
+  /**
+   * @summary The number of digits of precision in floating-point numbers.
+   * @constant {number}
+   */
+  m.digits = 53;
+  /**
+   * @summary The machine epsilon.
+   * @constant {number}
+   */
+  m.epsilon = Math.pow(0.5, m.digits - 1);
+  /**
+   * @summary The factor to convert degrees to radians.
+   * @constant {number}
+   */
+  m.degree = Math.PI/180;
+
+  /**
+   * @summary Square a number.
+   * @param {number} x the number.
+   * @return {number} the square.
+   */
+  m.sq = function(x) { return x * x; };
+
+  /**
+   * @summary The hypotenuse function.
+   * @param {number} x the first side.
+   * @param {number} y the second side.
+   * @return {number} the hypotenuse.
+   */
+  m.hypot = function(x, y) {
+    // Built in Math.hypot give incorrect results from GeodSolve92.
+    return Math.sqrt(x*x + y*y);
+  };
+
+  /**
+   * @summary Cube root function.
+   * @param {number} x the argument.
+   * @return {number} the real cube root.
+   */
+  m.cbrt = Math.cbrt || function(x) {
+    var y = Math.pow(Math.abs(x), 1/3);
+    return x > 0 ? y : (x < 0 ? -y : x);
+  };
+
+  /**
+   * @summary The log1p function.
+   * @param {number} x the argument.
+   * @return {number} log(1 + x).
+   */
+  m.log1p = Math.log1p || function(x) {
+    var y = 1 + x,
+        z = y - 1;
+    // Here's the explanation for this magic: y = 1 + z, exactly, and z
+    // approx x, thus log(y)/z (which is nearly constant near z = 0) returns
+    // a good approximation to the true log(1 + x)/x.  The multiplication x *
+    // (log(y)/z) introduces little additional error.
+    return z === 0 ? x : x * Math.log(y) / z;
+  };
+
+  /**
+   * @summary Inverse hyperbolic tangent.
+   * @param {number} x the argument.
+   * @return {number} tanh<sup>&minus;1</sup> x.
+   */
+  m.atanh = Math.atanh || function(x) {
+    var y = Math.abs(x);          // Enforce odd parity
+    y = m.log1p(2 * y/(1 - y))/2;
+    return x > 0 ? y : (x < 0 ? -y : x);
+  };
+
+  /**
+   * @summary Copy the sign.
+   * @param {number} x gives the magitude of the result.
+   * @param {number} y gives the sign of the result.
+   * @return {number} value with the magnitude of x and with the sign of y.
+   */
+  m.copysign = function(x, y) {
+    return Math.abs(x) * (y < 0 || (y === 0 && 1/y < 0) ? -1 : 1);
+  };
+
+  /**
+   * @summary An error-free sum.
+   * @param {number} u
+   * @param {number} v
+   * @return {object} sum with sum.s = round(u + v) and sum.t is u + v &minus;
+   *   round(u + v)
+   */
+  m.sum = function(u, v) {
+    var s = u + v,
+        up = s - v,
+        vpp = s - up,
+        t;
+    up -= u;
+    vpp -= v;
+    // if s = 0, then t = 0 and give t the same sign as s
+    t = s ? 0 - (up + vpp) : s;
+    // u + v =       s      + t
+    //       = round(u + v) + t
+    return {s: s, t: t};
+  };
+
+  /**
+   * @summary Evaluate a polynomial.
+   * @param {integer} N the order of the polynomial.
+   * @param {array} p the coefficient array (of size N + 1) (leading
+   *   order coefficient first)
+   * @param {number} x the variable.
+   * @return {number} the value of the polynomial.
+   */
+  m.polyval = function(N, p, s, x) {
+    var y = N < 0 ? 0 : p[s++];
+    while (--N >= 0) y = y * x + p[s++];
+    return y;
+  };
+
+  /**
+   * @summary Coarsen a value close to zero.
+   * @param {number} x
+   * @return {number} the coarsened value.
+   */
+  m.AngRound = function(x) {
+    // The makes the smallest gap in x = 1/16 - nextafter(1/16, 0) = 1/2^57 for
+    // reals = 0.7 pm on the earth if x is an angle in degrees.  (This is about
+    // 1000 times more resolution than we get with angles around 90 degrees.)
+    // We use this to avoid having to deal with near singular cases when x is
+    // non-zero but tiny (e.g., 1.0e-200).  This converts -0 to +0; however
+    // tiny negative numbers get converted to -0.
+    var z = 1/16,
+        y = Math.abs(x);
+    // The compiler mustn't "simplify" z - (z - y) to y
+    y = y < z ? z - (z - y) : y;
+    return m.copysign(y, x);
+  };
+
+  /**
+   * @summary The remainder function.
+   * @param {number} x the numerator of the division
+   * @param {number} y the denominator of the division
+   * @return {number} the remainder in the range [&minus;y/2, y/2].
+   * <p>
+   * The range of x is unrestricted; y must be positive.
+   */
+  m.remainder = function(x, y) {
+    x %= y;
+    return x < -y/2 ? x + y : (x < y/2 ? x : x - y);
+  };
+
+  /**
+   * @summary Normalize an angle.
+   * @param {number} x the angle in degrees.
+   * @return {number} the angle reduced to the range [&minus;180&deg;,
+   *   180&deg;].
+   *
+   * The range of x is unrestricted.  If the result is &plusmn;0&deg; or
+   * &plusmn;180&deg; then the sign is the sign of \e x.
+   */
+  m.AngNormalize = function(x) {
+    // Place angle in [-180, 180].
+    var y = m.remainder(x, 360);
+    return Math.abs(y) === 180 ? m.copysign(180, x) :  y;
+  };
+
+  /**
+   * @summary Normalize a latitude.
+   * @param {number} x the angle in degrees.
+   * @return {number} x if it is in the range [&minus;90&deg;, 90&deg;],
+   *   otherwise return NaN.
+   */
+  m.LatFix = function(x) {
+    // Replace angle with NaN if outside [-90, 90].
+    return Math.abs(x) > 90 ? NaN : x;
+  };
+
+  /**
+   * @summary The exact difference of two angles reduced to [&minus;180&deg;,
+   *   180&deg;]
+   * @param {number} x the first angle in degrees.
+   * @param {number} y the second angle in degrees.
+   * @return {object} diff the exact difference, diff.d + diff.e = y &minus; x
+   *   mod 360&deg;.
+   *
+   * This computes z = y &minus; x exactly, reduced to
+   * [&minus;180&deg;, 180&deg;]; and then sets z = d + e where d
+   * is the nearest representable number to z and e is the truncation
+   * error.  If z = &plusmn;0&deg; or &plusmn;180&deg;, then the sign of
+   * d is given by the sign of y &minus; x.  The maximum absolute
+   * value of e is 2<sup>&minus;26</sup> (for doubles).
+   */
+  m.AngDiff = function(x, y) {
+    // Compute y - x and reduce to [-180,180] accurately.
+    var r = m.sum(m.remainder(-x, 360), m.remainder(y, 360)), d, e;
+    r = m.sum(m.remainder(r.s, 360), r.t);
+    d = r.s;
+    e = r.t;
+    // Fix the sign if d = -180, 0, 180.
+    if (d === 0 || Math.abs(d) === 180)
+      // If e == 0, take sign from y - x
+      // else (e != 0, implies d = +/-180), d and e must have opposite signs
+      d = m.copysign(d, e === 0 ? y - x : -e);
+    return {d: d, e: e};
+  };
+
+  /**
+   * @summary Evaluate the sine and cosine function with the argument in
+   *   degrees
+   * @param {number} x in degrees.
+   * @return {object} r with r.s = sin(x) and r.c = cos(x).
+   */
+  m.sincosd = function(x) {
+    // In order to minimize round-off errors, this function exactly reduces
+    // the argument to the range [-45, 45] before converting it to radians.
+    var d, r, q, s, c, sinx, cosx;
+    d = x % 360;
+    q = Math.round(d / 90);     // If d is NaN this returns NaN
+    d -= 90 * q;
+    // now abs(d) <= 45
+    r = d * this.degree;
+    // Possibly could call the gnu extension sincos
+    s = Math.sin(r); c = Math.cos(r);
+    if (Math.abs(d) === 45) {
+      c = Math.sqrt(0.5);
+      s = m.copysign(c, r);
+    } else if (Math.abs(d) === 30) {
+      c = Math.sqrt(0.75);
+      s = m.copysign(0.5, r);
+    }
+    switch (q & 3) {
+    case 0:  sinx =  s; cosx =  c; break;
+    case 1:  sinx =  c; cosx = -s; break;
+    case 2:  sinx = -s; cosx = -c; break;
+    default: sinx = -c; cosx =  s; break; // case 3
+    }
+    cosx += 0;
+    if (sinx === 0) sinx = m.copysign(sinx, x);
+    return {s: sinx, c: cosx};
+  };
+
+  /**
+   * @summary Evaluate the sine and cosine with reduced argument plus correction
+   *
+   * @param {number} x reduced angle in degrees.
+   * @param {number} t correction in degrees.
+   * @return {object} r with r.s = sin(x + t) and r.c = cos(x + t).
+   */
+  m.sincosde = function(x, t) {
+    // In order to minimize round-off errors, this function exactly reduces
+    // the argument to the range [-45, 45] before converting it to radians.
+    var d, r, q, s, c, sinx, cosx;
+    d = x % 360;
+    q = Math.round(d / 90);     // If d is NaN this returns NaN
+    d = m.AngRound((d - 90 * q) + t);
+    // now abs(d) <= 45
+    r = d * this.degree;
+    // Possibly could call the gnu extension sincos
+    s = Math.sin(r); c = Math.cos(r);
+    if (Math.abs(d) === 45) {
+      c = Math.sqrt(0.5);
+      s = m.copysign(c, r);
+    } else if (Math.abs(d) === 30) {
+      c = Math.sqrt(0.75);
+      s = m.copysign(0.5, r);
+    }
+    switch (q & 3) {
+    case 0:  sinx =  s; cosx =  c; break;
+    case 1:  sinx =  c; cosx = -s; break;
+    case 2:  sinx = -s; cosx = -c; break;
+    default: sinx = -c; cosx =  s; break; // case 3
+    }
+    cosx += 0;
+    if (sinx === 0) sinx = m.copysign(sinx, x+t);
+    return {s: sinx, c: cosx};
+  };
+
+  /**
+   * @summary Evaluate the atan2 function with the result in degrees
+   * @param {number} y
+   * @param {number} x
+   * @return atan2(y, x) in degrees, in the range [&minus;180&deg;
+   *   180&deg;].
+   */
+  m.atan2d = function(y, x) {
+    // In order to minimize round-off errors, this function rearranges the
+    // arguments so that result of atan2 is in the range [-pi/4, pi/4] before
+    // converting it to degrees and mapping the result to the correct
+    // quadrant.
+    var q = 0, ang;
+    if (Math.abs(y) > Math.abs(x)) { [y, x] = [x, y]; q = 2; } // swap(x, y)
+    if (m.copysign(1, x) < 0) { x = -x; ++q; }
+    // here x >= 0 and x >= abs(y), so angle is in [-pi/4, pi/4]
+    ang = Math.atan2(y, x) / this.degree;
+    switch (q) {
+      // Note that atan2d(-0.0, 1.0) will return -0.  However, we expect that
+      // atan2d will not be called with y = -0.  If need be, include
+      //
+      //   case 0: ang = 0 + ang; break;
+      //
+      // and handle mpfr as in AngRound.
+    case 1: ang = m.copysign(180, y) - ang; break;
+    case 2: ang =             90     - ang; break;
+    case 3: ang =            -90     + ang; break;
+    default: break;
+    }
+    return ang;
+  };
+})(geodesic.Math);
+
+(function(
+  /**
+   * @exports geodesic/Accumulator
+   * @description Accurate summation via the
+   *   {@link module:geodesic/Accumulator.Accumulator Accumulator} class
+   *   (mainly for internal use).
+   */
+  a, m) {
+  "use strict";
+
+  /**
+   * @class
+   * @summary Accurate summation of many numbers.
+   * @classdesc This allows many numbers to be added together with twice the
+   *   normal precision.  In the documentation of the member functions, sum
+   *   stands for the value currently held in the accumulator.
+   * @param {number | Accumulator} [y = 0]  set sum = y.
+   */
+  a.Accumulator = function(y) {
+    this.Set(y);
+  };
+
+  /**
+   * @summary Set the accumulator to a number.
+   * @param {number | Accumulator} [y = 0] set sum = y.
+   */
+  a.Accumulator.prototype.Set = function(y) {
+    if (!y) y = 0;
+    if (y.constructor === a.Accumulator) {
+      this._s = y._s;
+      this._t = y._t;
+    } else {
+      this._s = y;
+      this._t = 0;
+    }
+  };
+
+  /**
+   * @summary Add a number to the accumulator.
+   * @param {number} [y = 0] set sum += y.
+   */
+  a.Accumulator.prototype.Add = function(y) {
+    // Here's Shewchuk's solution...
+    // Accumulate starting at least significant end
+    var u = m.sum(y, this._t),
+        v = m.sum(u.s, this._s);
+    u = u.t;
+    this._s = v.s;
+    this._t = v.t;
+    // Start is _s, _t decreasing and non-adjacent.  Sum is now (s + t + u)
+    // exactly with s, t, u non-adjacent and in decreasing order (except
+    // for possible zeros).  The following code tries to normalize the
+    // result.  Ideally, we want _s = round(s+t+u) and _u = round(s+t+u -
+    // _s).  The follow does an approximate job (and maintains the
+    // decreasing non-adjacent property).  Here are two "failures" using
+    // 3-bit floats:
+    //
+    // Case 1: _s is not equal to round(s+t+u) -- off by 1 ulp
+    // [12, -1] - 8 -> [4, 0, -1] -> [4, -1] = 3 should be [3, 0] = 3
+    //
+    // Case 2: _s+_t is not as close to s+t+u as it shold be
+    // [64, 5] + 4 -> [64, 8, 1] -> [64,  8] = 72 (off by 1)
+    //                    should be [80, -7] = 73 (exact)
+    //
+    // "Fixing" these problems is probably not worth the expense.  The
+    // representation inevitably leads to small errors in the accumulated
+    // values.  The additional errors illustrated here amount to 1 ulp of
+    // the less significant word during each addition to the Accumulator
+    // and an additional possible error of 1 ulp in the reported sum.
+    //
+    // Incidentally, the "ideal" representation described above is not
+    // canonical, because _s = round(_s + _t) may not be true.  For
+    // example, with 3-bit floats:
+    //
+    // [128, 16] + 1 -> [160, -16] -- 160 = round(145).
+    // But [160, 0] - 16 -> [128, 16] -- 128 = round(144).
+    //
+    if (this._s === 0)          // This implies t == 0,
+      this._s = u;              // so result is u
+    else
+      this._t += u;             // otherwise just accumulate u to t.
+  };
+
+  /**
+   * @summary Return the result of adding a number to sum (but
+   *   don't change sum).
+   * @param {number} [y = 0] the number to be added to the sum.
+   * @return sum + y.
+   */
+  a.Accumulator.prototype.Sum = function(y) {
+    var b;
+    if (!y)
+      return this._s;
+    else {
+      b = new a.Accumulator(this);
+      b.Add(y);
+      return b._s;
+    }
+  };
+
+  /**
+   * @summary Set sum = &minus;sum.
+   */
+  a.Accumulator.prototype.Negate = function() {
+    this._s *= -1;
+    this._t *= -1;
+  };
+
+  /**
+   * @summary Take the remainder
+   * @param {number} y the divisor of the remainder operation.
+   * @return sum in range [&minus;y/2, y/2].
+   */
+  a.Accumulator.prototype.Remainder = function(y) {
+    this._s = m.remainder(this._s, y);
+    this.Add(0);
+  };
+})(geodesic.Accumulator, geodesic.Math);
+
+/*
+ * Geodesic.js
+ * Transcription of Geodesic.[ch]pp into JavaScript.
+ *
+ * See the documentation for the C++ class.  The conversion is a literal
+ * conversion from C++.
+ *
+ * The algorithms are derived in
+ *
+ *    Charles F. F. Karney,
+ *    Algorithms for geodesics, J. Geodesy 87, 43-55 (2013);
+ *    https://doi.org/10.1007/s00190-012-0578-z
+ *    Addenda: https://geographiclib.sourceforge.io/geod-addenda.html
+ *
+ * Copyright (c) Charles Karney (2011-2022) <karney@alum.mit.edu> and licensed
+ * under the MIT/X11 License.  For more information, see
+ * https://geographiclib.sourceforge.io/
+ */
+
+// Load AFTER Math.js
+
+// To allow swap via [y, x] = [x, y]
+/* jshint esversion: 6 */
+
+geodesic.Geodesic = {};
+geodesic.GeodesicLine = {};
+geodesic.PolygonArea = {};
+
+(function(
+  /**
+   * @exports geodesic/Geodesic
+   * @description Solve geodesic problems via the
+   *   {@link module:geodesic/Geodesic.Geodesic Geodesic} class.
+   */
+  g, l, p, m, c) {
+  "use strict";
+
+  var GEOGRAPHICLIB_GEODESIC_ORDER = 6,
+      nA1_ = GEOGRAPHICLIB_GEODESIC_ORDER,
+      nA2_ = GEOGRAPHICLIB_GEODESIC_ORDER,
+      nA3_ = GEOGRAPHICLIB_GEODESIC_ORDER,
+      nA3x_ = nA3_,
+      nC3x_, nC4x_,
+      maxit1_ = 20,
+      maxit2_ = maxit1_ + m.digits + 10,
+      tol0_ = m.epsilon,
+      tol1_ = 200 * tol0_,
+      tol2_ = Math.sqrt(tol0_),
+      tolb_ = tol0_,
+      xthresh_ = 1000 * tol2_,
+      CAP_NONE = 0,
+      CAP_ALL  = 0x1F,
+      OUT_ALL  = 0x7F80,
+      astroid,
+      A1m1f_coeff, C1f_coeff, C1pf_coeff,
+      A2m1f_coeff, C2f_coeff,
+      A3_coeff, C3_coeff, C4_coeff;
+
+  // N.B. Number.MIN_VALUE is denormalized; divide by Number.EPSILON to get min
+  // normalized positive number.
+  g.tiny_ = Math.sqrt(Number.MIN_VALUE/Number.EPSILON);
+  g.nC1_ = GEOGRAPHICLIB_GEODESIC_ORDER;
+  g.nC1p_ = GEOGRAPHICLIB_GEODESIC_ORDER;
+  g.nC2_ = GEOGRAPHICLIB_GEODESIC_ORDER;
+  g.nC3_ = GEOGRAPHICLIB_GEODESIC_ORDER;
+  g.nC4_ = GEOGRAPHICLIB_GEODESIC_ORDER;
+  nC3x_ = (g.nC3_ * (g.nC3_ - 1)) / 2;
+  nC4x_ = (g.nC4_ * (g.nC4_ + 1)) / 2;
+  g.CAP_C1   = 1<<0;
+  g.CAP_C1p  = 1<<1;
+  g.CAP_C2   = 1<<2;
+  g.CAP_C3   = 1<<3;
+  g.CAP_C4   = 1<<4;
+
+  g.NONE          = 0;
+  g.ARC           = 1<<6;
+  g.LATITUDE      = 1<<7  | CAP_NONE;
+  g.LONGITUDE     = 1<<8  | g.CAP_C3;
+  g.AZIMUTH       = 1<<9  | CAP_NONE;
+  g.DISTANCE      = 1<<10 | g.CAP_C1;
+  g.STANDARD      = g.LATITUDE | g.LONGITUDE | g.AZIMUTH | g.DISTANCE;
+  g.DISTANCE_IN   = 1<<11 | g.CAP_C1 | g.CAP_C1p;
+  g.REDUCEDLENGTH = 1<<12 | g.CAP_C1 | g.CAP_C2;
+  g.GEODESICSCALE = 1<<13 | g.CAP_C1 | g.CAP_C2;
+  g.AREA          = 1<<14 | g.CAP_C4;
+  g.ALL           = OUT_ALL| CAP_ALL;
+  g.LONG_UNROLL   = 1<<15;
+  g.OUT_MASK      = OUT_ALL| g.LONG_UNROLL;
+
+  // Original math.gl stable difference: cos(k*(x+d))-cos(k*x)
+  // = -2*sin(k*(x+d/2))*sin(k*d/2), k odd. Retain tiny arc bits.
+  g.SinCosSeries = function(sinp, sinx, cosx, c, delta) {
+    if (delta !== undefined) {
+      var h = delta / 2, sh = Math.sin(h), ch = Math.cos(h);
+      // sin((2*i+1)*mid) = (-1)^i*cos((2*i+1)*(pi/2-mid)).
+      return -2 * g.SinCosSeries(false, cosx * ch - sinx * sh,
+        sinx * ch + cosx * sh,
+        c.map(function(v, i) {return (i % 2 ? -v : v) * Math.sin((2 * i + 1) * h);}));
+    }
+
+    // Evaluate
+    // y = sinp ? sum(c[i] * sin( 2*i    * x), i, 1, n) :
+    //            sum(c[i] * cos((2*i+1) * x), i, 0, n-1)
+    // using Clenshaw summation.  N.B. c[0] is unused for sin series
+    // Approx operation count = (n + 5) mult and (2 * n + 2) add
+    var k = c.length,           // Point to one beyond last element
+        n = k - (sinp ? 1 : 0),
+        ar = 2 * (cosx - sinx) * (cosx + sinx), // 2 * cos(2 * x)
+        y0 = n & 1 ? c[--k] : 0, y1 = 0;        // accumulators for sum
+    // Now n is even
+    n = Math.floor(n/2);
+    while (n--) {
+      // Unroll loop x 2, so accumulators return to their original role
+      y1 = ar * y0 - y1 + c[--k];
+      y0 = ar * y1 - y0 + c[--k];
+    }
+    return (sinp ? 2 * sinx * cosx * y0 : // sin(2 * x) * y0
+            cosx * (y0 - y1));            // cos(x) * (y0 - y1)
+  };
+
+  astroid = function(x, y) {
+    // Solve k^4+2*k^3-(x^2+y^2-1)*k^2-2*y^2*k-y^2 = 0 for positive
+    // root k.  This solution is adapted from Geocentric::Reverse.
+    var k,
+        p = m.sq(x),
+        q = m.sq(y),
+        r = (p + q - 1) / 6,
+        S, r2, r3, disc, u, T3, T, ang, v, uv, w;
+    if ( !(q === 0 && r <= 0) ) {
+      // Avoid possible division by zero when r = 0 by multiplying
+      // equations for s and t by r^3 and r, resp.
+      S = p * q / 4;            // S = r^3 * s
+      r2 = m.sq(r);
+      r3 = r * r2;
+      // The discriminant of the quadratic equation for T3.  This is
+      // zero on the evolute curve p^(1/3)+q^(1/3) = 1
+      disc = S * (S + 2 * r3);
+      u = r;
+      if (disc >= 0) {
+        T3 = S + r3;
+        // Pick the sign on the sqrt to maximize abs(T3).  This
+        // minimizes loss of precision due to cancellation.  The
+        // result is unchanged because of the way the T is used
+        // in definition of u.
+        T3 += T3 < 0 ? -Math.sqrt(disc) : Math.sqrt(disc);    // T3 = (r * t)^3
+        // N.B. cbrt always returns the real root.  cbrt(-8) = -2.
+        T = m.cbrt(T3);     // T = r * t
+        // T can be zero; but then r2 / T -> 0.
+        u += T + (T !== 0 ? r2 / T : 0);
+      } else {
+        // T is complex, but the way u is defined the result is real.
+        ang = Math.atan2(Math.sqrt(-disc), -(S + r3));
+        // There are three possible cube roots.  We choose the
+        // root which avoids cancellation.  Note that disc < 0
+        // implies that r < 0.
+        u += 2 * r * Math.cos(ang / 3);
+      }
+      v = Math.sqrt(m.sq(u) + q);       // guaranteed positive
+      // Avoid loss of accuracy when u < 0.
+      uv = u < 0 ? q / (v - u) : u + v; // u+v, guaranteed positive
+      w = (uv - q) / (2 * v);           // positive?
+      // Rearrange expression for k to avoid loss of accuracy due to
+      // subtraction.  Division by 0 not possible because uv > 0, w >= 0.
+      k = uv / (Math.sqrt(uv + m.sq(w)) + w); // guaranteed positive
+    } else {                                  // q == 0 && r <= 0
+      // y = 0 with |x| <= 1.  Handle this case directly.
+      // for y small, positive root is k = abs(y)/sqrt(1-x^2)
+      k = 0;
+    }
+    return k;
+  };
+
+  A1m1f_coeff = [
+    // (1-eps)*A1-1, polynomial in eps2 of order 3
+      +1, 4, 64, 0, 256
+  ];
+
+  // The scale factor A1-1 = mean value of (d/dsigma)I1 - 1
+  g.A1m1f = function(eps) {
+    var p = Math.floor(nA1_/2),
+        t = m.polyval(p, A1m1f_coeff, 0, m.sq(eps)) / A1m1f_coeff[p + 1];
+    return (t + eps) / (1 - eps);
+  };
+
+  C1f_coeff = [
+    // C1[1]/eps^1, polynomial in eps2 of order 2
+      -1, 6, -16, 32,
+    // C1[2]/eps^2, polynomial in eps2 of order 2
+      -9, 64, -128, 2048,
+    // C1[3]/eps^3, polynomial in eps2 of order 1
+      +9, -16, 768,
+    // C1[4]/eps^4, polynomial in eps2 of order 1
+      +3, -5, 512,
+    // C1[5]/eps^5, polynomial in eps2 of order 0
+      -7, 1280,
+    // C1[6]/eps^6, polynomial in eps2 of order 0
+      -7, 2048
+  ];
+
+  // The coefficients C1[l] in the Fourier expansion of B1
+  g.C1f = function(eps, c) {
+    var eps2 = m.sq(eps),
+        d = eps,
+        o = 0,
+        l, p;
+    for (l = 1; l <= g.nC1_; ++l) {     // l is index of C1p[l]
+      p = Math.floor((g.nC1_ - l) / 2); // order of polynomial in eps^2
+      c[l] = d * m.polyval(p, C1f_coeff, o, eps2) / C1f_coeff[o + p + 1];
+      o += p + 2;
+      d *= eps;
+    }
+  };
+
+  C1pf_coeff = [
+    // C1p[1]/eps^1, polynomial in eps2 of order 2
+      +205, -432, 768, 1536,
+    // C1p[2]/eps^2, polynomial in eps2 of order 2
+      +4005, -4736, 3840, 12288,
+    // C1p[3]/eps^3, polynomial in eps2 of order 1
+      -225, 116, 384,
+    // C1p[4]/eps^4, polynomial in eps2 of order 1
+      -7173, 2695, 7680,
+    // C1p[5]/eps^5, polynomial in eps2 of order 0
+      +3467, 7680,
+    // C1p[6]/eps^6, polynomial in eps2 of order 0
+      +38081, 61440
+  ];
+
+  // The coefficients C1p[l] in the Fourier expansion of B1p
+  g.C1pf = function(eps, c) {
+    var eps2 = m.sq(eps),
+        d = eps,
+        o = 0,
+        l, p;
+    for (l = 1; l <= g.nC1p_; ++l) {     // l is index of C1p[l]
+      p = Math.floor((g.nC1p_ - l) / 2); // order of polynomial in eps^2
+      c[l] = d * m.polyval(p, C1pf_coeff, o, eps2) / C1pf_coeff[o + p + 1];
+      o += p + 2;
+      d *= eps;
+    }
+  };
+
+  A2m1f_coeff = [
+    // (eps+1)*A2-1, polynomial in eps2 of order 3
+      -11, -28, -192, 0, 256
+  ];
+
+  // The scale factor A2-1 = mean value of (d/dsigma)I2 - 1
+  g.A2m1f = function(eps) {
+    var p = Math.floor(nA2_/2),
+        t = m.polyval(p, A2m1f_coeff, 0, m.sq(eps)) / A2m1f_coeff[p + 1];
+    return (t - eps) / (1 + eps);
+  };
+
+  C2f_coeff = [
+    // C2[1]/eps^1, polynomial in eps2 of order 2
+      +1, 2, 16, 32,
+    // C2[2]/eps^2, polynomial in eps2 of order 2
+      +35, 64, 384, 2048,
+    // C2[3]/eps^3, polynomial in eps2 of order 1
+      +15, 80, 768,
+    // C2[4]/eps^4, polynomial in eps2 of order 1
+      +7, 35, 512,
+    // C2[5]/eps^5, polynomial in eps2 of order 0
+      +63, 1280,
+    // C2[6]/eps^6, polynomial in eps2 of order 0
+      +77, 2048
+  ];
+
+  // The coefficients C2[l] in the Fourier expansion of B2
+  g.C2f = function(eps, c) {
+    var eps2 = m.sq(eps),
+        d = eps,
+        o = 0,
+        l, p;
+    for (l = 1; l <= g.nC2_; ++l) {     // l is index of C2[l]
+      p = Math.floor((g.nC2_ - l) / 2); // order of polynomial in eps^2
+      c[l] = d * m.polyval(p, C2f_coeff, o, eps2) / C2f_coeff[o + p + 1];
+      o += p + 2;
+      d *= eps;
+    }
+  };
+
+  /**
+   * @class
+   * @property {number} a the equatorial radius (meters).
+   * @property {number} f the flattening.
+   * @summary Initialize a Geodesic object for a specific ellipsoid.
+   * @classdesc Performs geodesic calculations on an ellipsoid of revolution.
+   *   The routines for solving the direct and inverse problems return an
+   *   object with some of the following fields set: lat1, lon1, azi1, lat2,
+   *   lon2, azi2, s12, a12, m12, M12, M21, S12.  See {@tutorial 2-interface},
+   *   section "The results".
+   * @example
+   * var geodesic = require("geographiclib-geodesic"),
+   *     geod = geodesic.Geodesic.WGS84;
+   * var inv = geod.Inverse(1,2,3,4);
+   * console.log("lat1 = " + inv.lat1 + ", lon1 = " + inv.lon1 +
+   *             ", lat2 = " + inv.lat2 + ", lon2 = " + inv.lon2 +
+   *             ",\nazi1 = " + inv.azi1 + ", azi2 = " + inv.azi2 +
+   *             ", s12 = " + inv.s12);
+   * @param {number} a the equatorial radius of the ellipsoid (meters).
+   * @param {number} f the flattening of the ellipsoid.  Setting f = 0 gives
+   *   a sphere (on which geodesics are great circles).  Negative f gives a
+   *   prolate ellipsoid.
+   * @throws an error if the parameters are illegal.
+   */
+  g.Geodesic = function(a, f) {
+    this.a = a;
+    this.f = f;
+    this._f1 = 1 - this.f;
+    this._e2 = this.f * (2 - this.f);
+    this._ep2 = this._e2 / m.sq(this._f1); // e2 / (1 - e2)
+    this._n = this.f / ( 2 - this.f);
+    this._b = this.a * this._f1;
+    // authalic radius squared
+    this._c2 = (m.sq(this.a) + m.sq(this._b) *
+                (this._e2 === 0 ? 1 :
+                 (this._e2 > 0 ? m.atanh(Math.sqrt(this._e2)) :
+                  Math.atan(Math.sqrt(-this._e2))) /
+                 Math.sqrt(Math.abs(this._e2))))/2;
+    // The sig12 threshold for "really short".  Using the auxiliary sphere
+    // solution with dnm computed at (bet1 + bet2) / 2, the relative error in
+    // the azimuth consistency check is sig12^2 * abs(f) * min(1, 1-f/2) / 2.
+    // (Error measured for 1/100 < b/a < 100 and abs(f) >= 1/1000.  For a given
+    // f and sig12, the max error occurs for lines near the pole.  If the old
+    // rule for computing dnm = (dn1 + dn2)/2 is used, then the error increases
+    // by a factor of 2.)  Setting this equal to epsilon gives sig12 = etol2.
+    // Here 0.1 is a safety factor (error decreased by 100) and max(0.001,
+    // abs(f)) stops etol2 getting too large in the nearly spherical case.
+    this._etol2 = 0.1 * tol2_ /
+      Math.sqrt( Math.max(0.001, Math.abs(this.f)) *
+                 Math.min(1, 1 - this.f/2) / 2 );
+    if (!(isFinite(this.a) && this.a > 0))
+      throw new Error("Equatorial radius is not positive");
+    if (!(isFinite(this._b) && this._b > 0))
+      throw new Error("Polar semi-axis is not positive");
+    this._A3x = new Array(nA3x_);
+    this._C3x = new Array(nC3x_);
+    this._C4x = new Array(nC4x_);
+    this.A3coeff();
+    this.C3coeff();
+    this.C4coeff();
+  };
+
+  A3_coeff = [
+    // A3, coeff of eps^5, polynomial in n of order 0
+      -3, 128,
+    // A3, coeff of eps^4, polynomial in n of order 1
+      -2, -3, 64,
+    // A3, coeff of eps^3, polynomial in n of order 2
+      -1, -3, -1, 16,
+    // A3, coeff of eps^2, polynomial in n of order 2
+      +3, -1, -2, 8,
+    // A3, coeff of eps^1, polynomial in n of order 1
+      +1, -1, 2,
+    // A3, coeff of eps^0, polynomial in n of order 0
+      +1, 1
+  ];
+
+  // The scale factor A3 = mean value of (d/dsigma)I3
+  g.Geodesic.prototype.A3coeff = function() {
+    var o = 0, k = 0,
+        j, p;
+    for (j = nA3_ - 1; j >= 0; --j) { // coeff of eps^j
+      p = Math.min(nA3_ - j - 1, j);  // order of polynomial in n
+      this._A3x[k++] = m.polyval(p, A3_coeff, o, this._n) /
+        A3_coeff[o + p + 1];
+      o += p + 2;
+    }
+  };
+
+  C3_coeff = [
+    // C3[1], coeff of eps^5, polynomial in n of order 0
+      +3, 128,
+    // C3[1], coeff of eps^4, polynomial in n of order 1
+      +2, 5, 128,
+    // C3[1], coeff of eps^3, polynomial in n of order 2
+      -1, 3, 3, 64,
+    // C3[1], coeff of eps^2, polynomial in n of order 2
+      -1, 0, 1, 8,
+    // C3[1], coeff of eps^1, polynomial in n of order 1
+      -1, 1, 4,
+    // C3[2], coeff of eps^5, polynomial in n of order 0
+      +5, 256,
+    // C3[2], coeff of eps^4, polynomial in n of order 1
+      +1, 3, 128,
+    // C3[2], coeff of eps^3, polynomial in n of order 2
+      -3, -2, 3, 64,
+    // C3[2], coeff of eps^2, polynomial in n of order 2
+      +1, -3, 2, 32,
+    // C3[3], coeff of eps^5, polynomial in n of order 0
+      +7, 512,
+    // C3[3], coeff of eps^4, polynomial in n of order 1
+      -10, 9, 384,
+    // C3[3], coeff of eps^3, polynomial in n of order 2
+      +5, -9, 5, 192,
+    // C3[4], coeff of eps^5, polynomial in n of order 0
+      +7, 512,
+    // C3[4], coeff of eps^4, polynomial in n of order 1
+      -14, 7, 512,
+    // C3[5], coeff of eps^5, polynomial in n of order 0
+      +21, 2560
+  ];
+
+  // The coefficients C3[l] in the Fourier expansion of B3
+  g.Geodesic.prototype.C3coeff = function() {
+    var o = 0, k = 0,
+        l, j, p;
+    for (l = 1; l < g.nC3_; ++l) {        // l is index of C3[l]
+      for (j = g.nC3_ - 1; j >= l; --j) { // coeff of eps^j
+        p = Math.min(g.nC3_ - j - 1, j);  // order of polynomial in n
+        this._C3x[k++] = m.polyval(p, C3_coeff, o, this._n) /
+          C3_coeff[o + p + 1];
+        o += p + 2;
+      }
+    }
+  };
+
+  C4_coeff = [
+    // C4[0], coeff of eps^5, polynomial in n of order 0
+      +97, 15015,
+    // C4[0], coeff of eps^4, polynomial in n of order 1
+      +1088, 156, 45045,
+    // C4[0], coeff of eps^3, polynomial in n of order 2
+      -224, -4784, 1573, 45045,
+    // C4[0], coeff of eps^2, polynomial in n of order 3
+      -10656, 14144, -4576, -858, 45045,
+    // C4[0], coeff of eps^1, polynomial in n of order 4
+      +64, 624, -4576, 6864, -3003, 15015,
+    // C4[0], coeff of eps^0, polynomial in n of order 5
+      +100, 208, 572, 3432, -12012, 30030, 45045,
+    // C4[1], coeff of eps^5, polynomial in n of order 0
+      +1, 9009,
+    // C4[1], coeff of eps^4, polynomial in n of order 1
+      -2944, 468, 135135,
+    // C4[1], coeff of eps^3, polynomial in n of order 2
+      +5792, 1040, -1287, 135135,
+    // C4[1], coeff of eps^2, polynomial in n of order 3
+      +5952, -11648, 9152, -2574, 135135,
+    // C4[1], coeff of eps^1, polynomial in n of order 4
+      -64, -624, 4576, -6864, 3003, 135135,
+    // C4[2], coeff of eps^5, polynomial in n of order 0
+      +8, 10725,
+    // C4[2], coeff of eps^4, polynomial in n of order 1
+      +1856, -936, 225225,
+    // C4[2], coeff of eps^3, polynomial in n of order 2
+      -8448, 4992, -1144, 225225,
+    // C4[2], coeff of eps^2, polynomial in n of order 3
+      -1440, 4160, -4576, 1716, 225225,
+    // C4[3], coeff of eps^5, polynomial in n of order 0
+      -136, 63063,
+    // C4[3], coeff of eps^4, polynomial in n of order 1
+      +1024, -208, 105105,
+    // C4[3], coeff of eps^3, polynomial in n of order 2
+      +3584, -3328, 1144, 315315,
+    // C4[4], coeff of eps^5, polynomial in n of order 0
+      -128, 135135,
+    // C4[4], coeff of eps^4, polynomial in n of order 1
+      -2560, 832, 405405,
+    // C4[5], coeff of eps^5, polynomial in n of order 0
+      +128, 99099
+  ];
+
+  g.Geodesic.prototype.C4coeff = function() {
+    var o = 0, k = 0,
+        l, j, p;
+    for (l = 0; l < g.nC4_; ++l) {        // l is index of C4[l]
+      for (j = g.nC4_ - 1; j >= l; --j) { // coeff of eps^j
+        p = g.nC4_ - j - 1;               // order of polynomial in n
+        this._C4x[k++] = m.polyval(p, C4_coeff, o, this._n) /
+          C4_coeff[o + p + 1];
+        o += p + 2;
+      }
+    }
+  };
+
+  g.Geodesic.prototype.A3f = function(eps) {
+    // Evaluate A3
+    return m.polyval(nA3x_ - 1, this._A3x, 0, eps);
+  };
+
+  g.Geodesic.prototype.C3f = function(eps, c) {
+    // Evaluate C3 coeffs
+    // Elements c[1] thru c[nC3_ - 1] are set
+    var mult = 1,
+        o = 0,
+        l, p;
+    for (l = 1; l < g.nC3_; ++l) { // l is index of C3[l]
+      p = g.nC3_ - l - 1;          // order of polynomial in eps
+      mult *= eps;
+      c[l] = mult * m.polyval(p, this._C3x, o, eps);
+      o += p + 1;
+    }
+  };
+
+  g.Geodesic.prototype.C4f = function(eps, c) {
+    // Evaluate C4 coeffs
+    // Elements c[0] thru c[g.nC4_ - 1] are set
+    var mult = 1,
+        o = 0,
+        l, p;
+    for (l = 0; l < g.nC4_; ++l) { // l is index of C4[l]
+      p = g.nC4_ - l - 1;          // order of polynomial in eps
+      c[l] = mult * m.polyval(p, this._C4x, o, eps);
+      o += p + 1;
+      mult *= eps;
+    }
+  };
+
+  // return s12b, m12b, m0, M12, M21
+  g.Geodesic.prototype.Lengths = function(eps, sig12,
+                                          ssig1, csig1, dn1, ssig2, csig2, dn2,
+                                          cbet1, cbet2, outmask,
+                                          C1a, C2a) {
+    // Return m12b = (reduced length)/_b; also calculate s12b =
+    // distance/_b, and m0 = coefficient of secular term in
+    // expression for reduced length.
+    outmask &= g.OUT_MASK;
+    var vals = {},
+        m0x = 0, J12 = 0, A1 = 0, A2 = 0,
+        B1, B2, l, csig12, t;
+    if (outmask & (g.DISTANCE | g.REDUCEDLENGTH | g.GEODESICSCALE)) {
+      A1 = g.A1m1f(eps);
+      g.C1f(eps, C1a);
+      if (outmask & (g.REDUCEDLENGTH | g.GEODESICSCALE)) {
+        A2 = g.A2m1f(eps);
+        g.C2f(eps, C2a);
+        m0x = A1 - A2;
+        A2 = 1 + A2;
+      }
+      A1 = 1 + A1;
+    }
+    if (outmask & g.DISTANCE) {
+      B1 = g.SinCosSeries(true, ssig2, csig2, C1a) -
+        g.SinCosSeries(true, ssig1, csig1, C1a);
+      // Missing a factor of _b
+      vals.s12b = A1 * (sig12 + B1);
+      if (outmask & (g.REDUCEDLENGTH | g.GEODESICSCALE)) {
+        B2 = g.SinCosSeries(true, ssig2, csig2, C2a) -
+          g.SinCosSeries(true, ssig1, csig1, C2a);
+        J12 = m0x * sig12 + (A1 * B1 - A2 * B2);
+      }
+    } else if (outmask & (g.REDUCEDLENGTH | g.GEODESICSCALE)) {
+      // Assume here that nC1_ >= nC2_
+      for (l = 1; l <= g.nC2_; ++l)
+        C2a[l] = A1 * C1a[l] - A2 * C2a[l];
+      J12 = m0x * sig12 + (g.SinCosSeries(true, ssig2, csig2, C2a) -
+                           g.SinCosSeries(true, ssig1, csig1, C2a));
+    }
+    if (outmask & g.REDUCEDLENGTH) {
+      vals.m0 = m0x;
+      // Missing a factor of _b.
+      // Add parens around (csig1 * ssig2) and (ssig1 * csig2) to ensure
+      // accurate cancellation in the case of coincident points.
+      vals.m12b = dn2 * (csig1 * ssig2) - dn1 * (ssig1 * csig2) -
+        csig1 * csig2 * J12;
+    } else
+      vals.m12b = NaN;
+    if (outmask & g.GEODESICSCALE) {
+      csig12 = csig1 * csig2 + ssig1 * ssig2;
+      t = this._ep2 * (cbet1 - cbet2) * (cbet1 + cbet2) / (dn1 + dn2);
+      vals.M12 = csig12 + (t * ssig2 - csig2 * J12) * ssig1 / dn1;
+      vals.M21 = csig12 - (t * ssig1 - csig1 * J12) * ssig2 / dn2;
+    }
+    return vals;
+  };
+
+  // return sig12, salp1, calp1, salp2, calp2, dnm
+  g.Geodesic.prototype.InverseStart = function(sbet1, cbet1, dn1,
+                                               sbet2, cbet2, dn2,
+                                               lam12, slam12, clam12,
+                                               C1a, C2a) {
+    // Return a starting point for Newton's method in salp1 and calp1
+    // (function value is -1).  If Newton's method doesn't need to be
+    // used, return also salp2 and calp2 and function value is sig12.
+    // salp2, calp2 only updated if return val >= 0.
+    var vals = {},
+        // bet12 = bet2 - bet1 in [0, pi); bet12a = bet2 + bet1 in (-pi, 0]
+        sbet12 = sbet2 * cbet1 - cbet2 * sbet1,
+        cbet12 = cbet2 * cbet1 + sbet2 * sbet1,
+        sbet12a, shortline, omg12, sbetm2, somg12, comg12, t, ssig12, csig12,
+        x, y, lamscale, betscale, k2, eps, cbet12a, bet12a, m12b, m0, nvals,
+        k, omg12a, lam12x;
+    vals.sig12 = -1;        // Return value
+    // Volatile declaration needed to fix inverse cases
+    // 88.202499451857 0 -88.202499451857 179.981022032992859592
+    // 89.262080389218 0 -89.262080389218 179.992207982775375662
+    // 89.333123580033 0 -89.333123580032997687 179.99295812360148422
+    // which otherwise fail with g++ 4.4.4 x86 -O3
+    sbet12a = sbet2 * cbet1;
+    sbet12a += cbet2 * sbet1;
+
+    shortline = cbet12 >= 0 && sbet12 < 0.5 && cbet2 * lam12 < 0.5;
+    if (shortline) {
+      sbetm2 = m.sq(sbet1 + sbet2);
+      // sin((bet1+bet2)/2)^2
+      // =  (sbet1 + sbet2)^2 / ((sbet1 + sbet2)^2 + (cbet1 + cbet2)^2)
+      sbetm2 /= sbetm2 + m.sq(cbet1 + cbet2);
+      vals.dnm = Math.sqrt(1 + this._ep2 * sbetm2);
+      omg12 = lam12 / (this._f1 * vals.dnm);
+      somg12 = Math.sin(omg12); comg12 = Math.cos(omg12);
+    } else {
+      somg12 = slam12; comg12 = clam12;
+    }
+
+    vals.salp1 = cbet2 * somg12;
+    vals.calp1 = comg12 >= 0 ?
+      sbet12 + cbet2 * sbet1 * m.sq(somg12) / (1 + comg12) :
+      sbet12a - cbet2 * sbet1 * m.sq(somg12) / (1 - comg12);
+
+    ssig12 = m.hypot(vals.salp1, vals.calp1);
+    csig12 = sbet1 * sbet2 + cbet1 * cbet2 * comg12;
+    if (shortline && ssig12 < this._etol2) {
+      // really short lines
+      vals.salp2 = cbet1 * somg12;
+      vals.calp2 = sbet12 - cbet1 * sbet2 *
+        (comg12 >= 0 ? m.sq(somg12) / (1 + comg12) : 1 - comg12);
+      // norm(vals.salp2, vals.calp2);
+      t = m.hypot(vals.salp2, vals.calp2); vals.salp2 /= t; vals.calp2 /= t;
+      // Set return value
+      vals.sig12 = Math.atan2(ssig12, csig12);
+    } else if (Math.abs(this._n) > 0.1 || // Skip astroid calc if too eccentric
+               csig12 >= 0 ||
+               ssig12 >= 6 * Math.abs(this._n) * Math.PI * m.sq(cbet1)) {
+      // Nothing to do, zeroth order spherical approximation is OK
+    } else {
+      // Scale lam12 and bet2 to x, y coordinate system where antipodal
+      // point is at origin and singular point is at y = 0, x = -1.
+      lam12x = Math.atan2(-slam12, -clam12); // lam12 - pi
+      if (this.f >= 0) {       // In fact f == 0 does not get here
+        // x = dlong, y = dlat
+        k2 = m.sq(sbet1) * this._ep2;
+        eps = k2 / (2 * (1 + Math.sqrt(1 + k2)) + k2);
+        lamscale = this.f * cbet1 * this.A3f(eps) * Math.PI;
+        betscale = lamscale * cbet1;
+
+        x = lam12x / lamscale;
+        y = sbet12a / betscale;
+      } else {                  // f < 0
+        // x = dlat, y = dlong
+        cbet12a = cbet2 * cbet1 - sbet2 * sbet1;
+        bet12a = Math.atan2(sbet12a, cbet12a);
+        // In the case of lon12 = 180, this repeats a calculation made
+        // in Inverse.
+        nvals = this.Lengths(this._n, Math.PI + bet12a,
+                             sbet1, -cbet1, dn1, sbet2, cbet2, dn2,
+                             cbet1, cbet2, g.REDUCEDLENGTH, C1a, C2a);
+        m12b = nvals.m12b; m0 = nvals.m0;
+        x = -1 + m12b / (cbet1 * cbet2 * m0 * Math.PI);
+        betscale = x < -0.01 ? sbet12a / x :
+          -this.f * m.sq(cbet1) * Math.PI;
+        lamscale = betscale / cbet1;
+        y = lam12 / lamscale;
+      }
+
+      if (y > -tol1_ && x > -1 - xthresh_) {
+        // strip near cut
+        if (this.f >= 0) {
+          vals.salp1 = Math.min(1, -x);
+          vals.calp1 = -Math.sqrt(1 - m.sq(vals.salp1));
+        } else {
+          vals.calp1 = Math.max(x > -tol1_ ? 0 : -1, x);
+          vals.salp1 = Math.sqrt(1 - m.sq(vals.calp1));
+        }
+      } else {
+        // Estimate alp1, by solving the astroid problem.
+        //
+        // Could estimate alpha1 = theta + pi/2, directly, i.e.,
+        //   calp1 = y/k; salp1 = -x/(1+k);  for f >= 0
+        //   calp1 = x/(1+k); salp1 = -y/k;  for f < 0 (need to check)
+        //
+        // However, it's better to estimate omg12 from astroid and use
+        // spherical formula to compute alp1.  This reduces the mean number of
+        // Newton iterations for astroid cases from 2.24 (min 0, max 6) to 2.12
+        // (min 0 max 5).  The changes in the number of iterations are as
+        // follows:
+        //
+        // change percent
+        //    1       5
+        //    0      78
+        //   -1      16
+        //   -2       0.6
+        //   -3       0.04
+        //   -4       0.002
+        //
+        // The histogram of iterations is (m = number of iterations estimating
+        // alp1 directly, n = number of iterations estimating via omg12, total
+        // number of trials = 148605):
+        //
+        //  iter    m      n
+        //    0   148    186
+        //    1 13046  13845
+        //    2 93315 102225
+        //    3 36189  32341
+        //    4  5396      7
+        //    5   455      1
+        //    6    56      0
+        //
+        // Because omg12 is near pi, estimate work with omg12a = pi - omg12
+        k = astroid(x, y);
+        omg12a = lamscale * ( this.f >= 0 ? -x * k/(1 + k) : -y * (1 + k)/k );
+        somg12 = Math.sin(omg12a); comg12 = -Math.cos(omg12a);
+        // Update spherical estimate of alp1 using omg12 instead of
+        // lam12
+        vals.salp1 = cbet2 * somg12;
+        vals.calp1 = sbet12a -
+          cbet2 * sbet1 * m.sq(somg12) / (1 - comg12);
+      }
+    }
+    // Sanity check on starting guess.  Backwards check allows NaN through.
+    // jshint -W018
+    if (!(vals.salp1 <= 0)) {
+      // norm(vals.salp1, vals.calp1);
+      t = m.hypot(vals.salp1, vals.calp1); vals.salp1 /= t; vals.calp1 /= t;
+    } else {
+      vals.salp1 = 1; vals.calp1 = 0;
+    }
+    return vals;
+  };
+
+  // return lam12, salp2, calp2, sig12, ssig1, csig1, ssig2, csig2, eps,
+  // domg12, dlam12,
+  g.Geodesic.prototype.Lambda12 = function(sbet1, cbet1, dn1,
+                                           sbet2, cbet2, dn2,
+                                           salp1, calp1, slam120, clam120,
+                                           diffp, C1a, C2a, C3a) {
+    var vals = {},
+        t, salp0, calp0,
+        somg1, comg1, somg2, comg2, somg12, comg12, B312, eta, k2, nvals;
+    if (sbet1 === 0 && calp1 === 0)
+      // Break degeneracy of equatorial line.  This case has already been
+      // handled.
+      calp1 = -g.tiny_;
+
+    // sin(alp1) * cos(bet1) = sin(alp0)
+    salp0 = salp1 * cbet1;
+    calp0 = m.hypot(calp1, salp1 * sbet1); // calp0 > 0
+
+    // tan(bet1) = tan(sig1) * cos(alp1)
+    // tan(omg1) = sin(alp0) * tan(sig1) = tan(omg1)=tan(alp1)*sin(bet1)
+    vals.ssig1 = sbet1; somg1 = salp0 * sbet1;
+    vals.csig1 = comg1 = calp1 * cbet1;
+    // norm(vals.ssig1, vals.csig1);
+    t = m.hypot(vals.ssig1, vals.csig1); vals.ssig1 /= t; vals.csig1 /= t;
+    // norm(somg1, comg1); -- don't need to normalize!
+
+    // Enforce symmetries in the case abs(bet2) = -bet1.  Need to be careful
+    // about this case, since this can yield singularities in the Newton
+    // iteration.
+    // sin(alp2) * cos(bet2) = sin(alp0)
+    vals.salp2 = cbet2 !== cbet1 ? salp0 / cbet2 : salp1;
+    // calp2 = sqrt(1 - sq(salp2))
+    //       = sqrt(sq(calp0) - sq(sbet2)) / cbet2
+    // and subst for calp0 and rearrange to give (choose positive sqrt
+    // to give alp2 in [0, pi/2]).
+    vals.calp2 = cbet2 !== cbet1 || Math.abs(sbet2) !== -sbet1 ?
+      Math.sqrt(m.sq(calp1 * cbet1) + (cbet1 < -sbet1 ?
+                                       (cbet2 - cbet1) * (cbet1 + cbet2) :
+                                       (sbet1 - sbet2) * (sbet1 + sbet2))) /
+      cbet2 : Math.abs(calp1);
+    // tan(bet2) = tan(sig2) * cos(alp2)
+    // tan(omg2) = sin(alp0) * tan(sig2).
+    vals.ssig2 = sbet2; somg2 = salp0 * sbet2;
+    vals.csig2 = comg2 = vals.calp2 * cbet2;
+    // norm(vals.ssig2, vals.csig2);
+    t = m.hypot(vals.ssig2, vals.csig2); vals.ssig2 /= t; vals.csig2 /= t;
+    // norm(somg2, comg2); -- don't need to normalize!
+
+    // sig12 = sig2 - sig1, limit to [0, pi]
+    vals.sig12 = Math.atan2(Math.max(0, vals.csig1 * vals.ssig2 -
+                                        vals.ssig1 * vals.csig2),
+                                        vals.csig1 * vals.csig2 +
+                                        vals.ssig1 * vals.ssig2);
+
+    // omg12 = omg2 - omg1, limit to [0, pi]
+    somg12 = Math.max(0, comg1 * somg2 - somg1 * comg2);
+    comg12 =             comg1 * comg2 + somg1 * somg2;
+    // eta = omg12 - lam120
+    eta = Math.atan2(somg12 * clam120 - comg12 * slam120,
+                     comg12 * clam120 + somg12 * slam120);
+    k2 = m.sq(calp0) * this._ep2;
+    vals.eps = k2 / (2 * (1 + Math.sqrt(1 + k2)) + k2);
+    this.C3f(vals.eps, C3a);
+    B312 = (g.SinCosSeries(true, vals.ssig2, vals.csig2, C3a) -
+            g.SinCosSeries(true, vals.ssig1, vals.csig1, C3a));
+    vals.domg12 =  -this.f * this.A3f(vals.eps) * salp0 * (vals.sig12 + B312);
+    vals.lam12 = eta + vals.domg12;
+    if (diffp) {
+      if (vals.calp2 === 0)
+        vals.dlam12 = -2 * this._f1 * dn1 / sbet1;
+      else {
+        nvals = this.Lengths(vals.eps, vals.sig12,
+                             vals.ssig1, vals.csig1, dn1,
+                             vals.ssig2, vals.csig2, dn2,
+                             cbet1, cbet2, g.REDUCEDLENGTH, C1a, C2a);
+        vals.dlam12 = nvals.m12b;
+        vals.dlam12 *= this._f1 / (vals.calp2 * cbet2);
+      }
+    }
+    return vals;
+  };
+
+  /**
+   * @summary Solve the inverse geodesic problem.
+   * @param {number} lat1 the latitude of the first point in degrees.
+   * @param {number} lon1 the longitude of the first point in degrees.
+   * @param {number} lat2 the latitude of the second point in degrees.
+   * @param {number} lon2 the longitude of the second point in degrees.
+   * @param {bitmask} [outmask = STANDARD] which results to include.
+   * @return {object} the requested results
+   * @description The lat1, lon1, lat2, lon2, and a12 fields of the result are
+   *   always set.  For details on the outmask parameter, see {@tutorial
+   *   2-interface}, "The outmask and caps parameters".
+   */
+  g.Geodesic.prototype.Inverse = function(lat1, lon1, lat2, lon2, outmask) {
+    var r, vals;
+    if (!outmask) outmask = g.STANDARD;
+    if (outmask === g.LONG_UNROLL) outmask |= g.STANDARD;
+    outmask &= g.OUT_MASK;
+    r = this.InverseInt(lat1, lon1, lat2, lon2, outmask);
+    vals = r.vals;
+    if (outmask & g.AZIMUTH) {
+      vals.azi1 = m.atan2d(r.salp1, r.calp1);
+      vals.azi2 = m.atan2d(r.salp2, r.calp2);
+    }
+    return vals;
+  };
+
+  g.Geodesic.prototype.InverseInt = function(lat1, lon1, lat2, lon2, outmask) {
+    var vals = {},
+        lon12, lon12s, lonsign, t, swapp, latsign,
+        sbet1, cbet1, sbet2, cbet2, s12x, m12x,
+        dn1, dn2, lam12, slam12, clam12,
+        sig12, calp1, salp1, calp2, salp2, C1a, C2a, C3a, meridian, nvals,
+        ssig1, csig1, ssig2, csig2, eps, omg12, dnm,
+        numit, salp1a, calp1a, salp1b, calp1b,
+        tripn, tripb, v, dv, dalp1, sdalp1, cdalp1, nsalp1,
+        lengthmask, salp0, calp0, alp12, k2, A4, C4a, B41, B42,
+        somg12, comg12, domg12, dbet1, dbet2, salp12, calp12, sdomg12, cdomg12;
+    // Compute longitude difference (AngDiff does this carefully).  Result is
+    // in [-180, 180] but -180 is only for west-going geodesics.  180 is for
+    // east-going and meridional geodesics.
+    vals.lat1 = lat1 = m.LatFix(lat1); vals.lat2 = lat2 = m.LatFix(lat2);
+    // If really close to the equator, treat as on equator.
+    lat1 = m.AngRound(lat1);
+    lat2 = m.AngRound(lat2);
+    lon12 = m.AngDiff(lon1, lon2); lon12s = lon12.e; lon12 = lon12.d;
+    if (outmask & g.LONG_UNROLL) {
+      vals.lon1 = lon1; vals.lon2 = (lon1 + lon12) + lon12s;
+    } else {
+      vals.lon1 = m.AngNormalize(lon1); vals.lon2 = m.AngNormalize(lon2);
+    }
+    // Make longitude difference positive.
+    lonsign = m.copysign(1, lon12);
+    lon12 *= lonsign; lon12s *= lonsign;
+    lam12 = lon12 * m.degree;
+    // Calculate sincos of lon12 + error (this applies AngRound internally).
+    t = m.sincosde(lon12, lon12s); slam12 = t.s; clam12 = t.c;
+    lon12s = (180 - lon12) - lon12s; // the supplementary longitude difference
+
+    // Swap points so that point with higher (abs) latitude is point 1
+    // If one latitude is a nan, then it becomes lat1.
+    swapp = Math.abs(lat1) < Math.abs(lat2) || isNaN(lat2) ? -1 : 1;
+    if (swapp < 0) {
+      lonsign *= -1;
+      [lat2, lat1] = [lat1, lat2]; // swap(lat1, lat2);
+    }
+    // Make lat1 <= 0
+    latsign = m.copysign(1, -lat1);
+    lat1 *= latsign;
+    lat2 *= latsign;
+    // Now we have
+    //
+    //     0 <= lon12 <= 180
+    //     -90 <= lat1 <= 0
+    //     lat1 <= lat2 <= -lat1
+    //
+    // longsign, swapp, latsign register the transformation to bring the
+    // coordinates to this canonical form.  In all cases, 1 means no change was
+    // made.  We make these transformations so that there are few cases to
+    // check, e.g., on verifying quadrants in atan2.  In addition, this
+    // enforces some symmetries in the results returned.
+
+    t = m.sincosd(lat1); sbet1 = this._f1 * t.s; cbet1 = t.c;
+    // norm(sbet1, cbet1);
+    t = m.hypot(sbet1, cbet1); sbet1 /= t; cbet1 /= t;
+    // Ensure cbet1 = +epsilon at poles
+    cbet1 = Math.max(g.tiny_, cbet1);
+
+    t = m.sincosd(lat2); sbet2 = this._f1 * t.s; cbet2 = t.c;
+    // norm(sbet2, cbet2);
+    t = m.hypot(sbet2, cbet2); sbet2 /= t; cbet2 /= t;
+    // Ensure cbet2 = +epsilon at poles
+    cbet2 = Math.max(g.tiny_, cbet2);
+
+    // If cbet1 < -sbet1, then cbet2 - cbet1 is a sensitive measure of the
+    // |bet1| - |bet2|.  Alternatively (cbet1 >= -sbet1), abs(sbet2) + sbet1 is
+    // a better measure.  This logic is used in assigning calp2 in Lambda12.
+    // Sometimes these quantities vanish and in that case we force bet2 = +/-
+    // bet1 exactly.  An example where is is necessary is the inverse problem
+    // 48.522876735459 0 -48.52287673545898293 179.599720456223079643
+    // which failed with Visual Studio 10 (Release and Debug)
+
+    if (cbet1 < -sbet1) {
+      if (cbet2 === cbet1)
+        sbet2 = m.copysign(sbet1, sbet2);
+    } else {
+      if (Math.abs(sbet2) === -sbet1)
+        cbet2 = cbet1;
+    }
+
+    dn1 = Math.sqrt(1 + this._ep2 * m.sq(sbet1));
+    dn2 = Math.sqrt(1 + this._ep2 * m.sq(sbet2));
+
+    // index zero elements of these arrays are unused
+    C1a = new Array(g.nC1_ + 1);
+    C2a = new Array(g.nC2_ + 1);
+    C3a = new Array(g.nC3_);
+
+    meridian = lat1 === -90 || slam12 === 0;
+    if (meridian) {
+
+      // Endpoints are on a single full meridian, so the geodesic might
+      // lie on a meridian.
+
+      calp1 = clam12; salp1 = slam12; // Head to the target longitude
+      calp2 = 1; salp2 = 0;           // At the target we're heading north
+
+      // tan(bet) = tan(sig) * cos(alp)
+      ssig1 = sbet1; csig1 = calp1 * cbet1;
+      ssig2 = sbet2; csig2 = calp2 * cbet2;
+
+      // sig12 = sig2 - sig1
+      sig12 = Math.atan2(Math.max(0, csig1 * ssig2 - ssig1 * csig2),
+                                     csig1 * csig2 + ssig1 * ssig2);
+      nvals = this.Lengths(this._n, sig12,
+                           ssig1, csig1, dn1, ssig2, csig2, dn2, cbet1, cbet2,
+                           outmask | g.DISTANCE | g.REDUCEDLENGTH,
+                           C1a, C2a);
+      s12x = nvals.s12b;
+      m12x = nvals.m12b;
+      // Ignore m0
+      if (outmask & g.GEODESICSCALE) {
+        vals.M12 = nvals.M12;
+        vals.M21 = nvals.M21;
+      }
+      // Add the check for sig12 since zero length geodesics might yield
+      // m12 < 0.  Test case was
+      //
+      //    echo 20.001 0 20.001 0 | GeodSolve -i
+      if (sig12 < tol2_ || m12x >= 0) {
+        // Need at least 2, to handle 90 0 90 180
+        if (sig12 < 3 * g.tiny_ ||
+            // Prevent negative s12 or m12 for short lines
+            (sig12 < tol0_ && (s12x < 0 || m12x < 0)))
+          sig12 = m12x = s12x = 0;
+        m12x *= this._b;
+        s12x *= this._b;
+        vals.a12 = sig12 / m.degree;
+      } else
+        // m12 < 0, i.e., prolate and too close to anti-podal
+        meridian = false;
+    }
+
+    somg12 = 2;
+    if (!meridian &&
+        sbet1 === 0 &&           // and sbet2 == 0
+        (this.f <= 0 || lon12s >= this.f * 180)) {
+
+      // Geodesic runs along equator
+      calp1 = calp2 = 0; salp1 = salp2 = 1;
+      s12x = this.a * lam12;
+      sig12 = omg12 = lam12 / this._f1;
+      m12x = this._b * Math.sin(sig12);
+      if (outmask & g.GEODESICSCALE)
+        vals.M12 = vals.M21 = Math.cos(sig12);
+      vals.a12 = lon12 / this._f1;
+
+    } else if (!meridian) {
+
+      // Now point1 and point2 belong within a hemisphere bounded by a
+      // meridian and geodesic is neither meridional or equatorial.
+
+      // Figure a starting point for Newton's method
+      nvals = this.InverseStart(sbet1, cbet1, dn1, sbet2, cbet2, dn2,
+                                lam12, slam12, clam12, C1a, C2a);
+      sig12 = nvals.sig12;
+      salp1 = nvals.salp1;
+      calp1 = nvals.calp1;
+
+      if (sig12 >= 0) {
+        salp2 = nvals.salp2;
+        calp2 = nvals.calp2;
+        // Short lines (InverseStart sets salp2, calp2, dnm)
+
+        dnm = nvals.dnm;
+        s12x = sig12 * this._b * dnm;
+        m12x = m.sq(dnm) * this._b * Math.sin(sig12 / dnm);
+        if (outmask & g.GEODESICSCALE)
+          vals.M12 = vals.M21 = Math.cos(sig12 / dnm);
+        vals.a12 = sig12 / m.degree;
+        omg12 = lam12 / (this._f1 * dnm);
+      } else {
+
+        // Newton's method.  This is a straightforward solution of f(alp1) =
+        // lambda12(alp1) - lam12 = 0 with one wrinkle.  f(alp) has exactly one
+        // root in the interval (0, pi) and its derivative is positive at the
+        // root.  Thus f(alp) is positive for alp > alp1 and negative for alp <
+        // alp1.  During the course of the iteration, a range (alp1a, alp1b) is
+        // maintained which brackets the root and with each evaluation of
+        // f(alp) the range is shrunk if possible.  Newton's method is
+        // restarted whenever the derivative of f is negative (because the new
+        // value of alp1 is then further from the solution) or if the new
+        // estimate of alp1 lies outside (0,pi); in this case, the new starting
+        // guess is taken to be (alp1a + alp1b) / 2.
+        numit = 0;
+        // Bracketing range
+        salp1a = g.tiny_; calp1a = 1; salp1b = g.tiny_; calp1b = -1;
+        for (tripn = false, tripb = false;; ++numit) {
+          // the WGS84 test set: mean = 1.47, sd = 1.25, max = 16
+          // WGS84 and random input: mean = 2.85, sd = 0.60
+          nvals = this.Lambda12(sbet1, cbet1, dn1, sbet2, cbet2, dn2,
+                                salp1, calp1, slam12, clam12, numit < maxit1_,
+                                C1a, C2a, C3a);
+          v = nvals.lam12;
+          salp2 = nvals.salp2;
+          calp2 = nvals.calp2;
+          sig12 = nvals.sig12;
+          ssig1 = nvals.ssig1;
+          csig1 = nvals.csig1;
+          ssig2 = nvals.ssig2;
+          csig2 = nvals.csig2;
+          eps = nvals.eps;
+          domg12 = nvals.domg12;
+          dv = nvals.dlam12;
+
+          // Reversed test to allow escape with NaNs
+          // jshint -W018
+          if (tripb || !(Math.abs(v) >= (tripn ? 8 : 1) * tol0_) ||
+             numit == maxit2_)
+            break;
+          // Update bracketing values
+          if (v > 0 && (numit < maxit1_ || calp1/salp1 > calp1b/salp1b)) {
+            salp1b = salp1; calp1b = calp1;
+          } else if (v < 0 &&
+                     (numit < maxit1_ || calp1/salp1 < calp1a/salp1a)) {
+            salp1a = salp1; calp1a = calp1;
+          }
+          if (numit < maxit1_ && dv > 0) {
+            dalp1 = -v/dv;
+            if (Math.abs(dalp1) < Math.PI) {
+              sdalp1 = Math.sin(dalp1); cdalp1 = Math.cos(dalp1);
+              nsalp1 = salp1 * cdalp1 + calp1 * sdalp1;
+              if (nsalp1 > 0) {
+                calp1 = calp1 * cdalp1 - salp1 * sdalp1;
+                salp1 = nsalp1;
+                // norm(salp1, calp1);
+                t = m.hypot(salp1, calp1); salp1 /= t; calp1 /= t;
+                // In some regimes we don't get quadratic convergence because
+                // slope -> 0.  So use convergence conditions based on epsilon
+                // instead of sqrt(epsilon).
+                tripn = Math.abs(v) <= 16 * tol0_;
+                continue;
+              }
+            }
+          }
+          // Either dv was not positive or updated value was outside legal
+          // range.  Use the midpoint of the bracket as the next estimate.
+          // This mechanism is not needed for the WGS84 ellipsoid, but it does
+          // catch problems with more eccentric ellipsoids.  Its efficacy is
+          // such for the WGS84 test set with the starting guess set to alp1 =
+          // 90deg:
+          // the WGS84 test set: mean = 5.21, sd = 3.93, max = 24
+          // WGS84 and random input: mean = 4.74, sd = 0.99
+          salp1 = (salp1a + salp1b)/2;
+          calp1 = (calp1a + calp1b)/2;
+          // norm(salp1, calp1);
+          t = m.hypot(salp1, calp1); salp1 /= t; calp1 /= t;
+          tripn = false;
+          tripb = (Math.abs(salp1a - salp1) + (calp1a - calp1) < tolb_ ||
+                   Math.abs(salp1 - salp1b) + (calp1 - calp1b) < tolb_);
+        }
+        lengthmask = outmask |
+            (outmask & (g.REDUCEDLENGTH | g.GEODESICSCALE) ?
+             g.DISTANCE : g.NONE);
+        nvals = this.Lengths(eps, sig12,
+                             ssig1, csig1, dn1, ssig2, csig2, dn2,
+                             cbet1, cbet2,
+                             lengthmask, C1a, C2a);
+        s12x = nvals.s12b;
+        m12x = nvals.m12b;
+        // Ignore m0
+        if (outmask & g.GEODESICSCALE) {
+          vals.M12 = nvals.M12;
+          vals.M21 = nvals.M21;
+        }
+        m12x *= this._b;
+        s12x *= this._b;
+        vals.a12 = sig12 / m.degree;
+        if (outmask & g.AREA) {
+          // omg12 = lam12 - domg12
+          sdomg12 = Math.sin(domg12); cdomg12 = Math.cos(domg12);
+          somg12 = slam12 * cdomg12 - clam12 * sdomg12;
+          comg12 = clam12 * cdomg12 + slam12 * sdomg12;
+        }
+      }
+    }
+
+    if (outmask & g.DISTANCE)
+      vals.s12 = 0 + s12x;      // Convert -0 to 0
+
+    if (outmask & g.REDUCEDLENGTH)
+      vals.m12 = 0 + m12x;      // Convert -0 to 0
+
+    if (outmask & g.AREA) {
+      // From Lambda12: sin(alp1) * cos(bet1) = sin(alp0)
+      salp0 = salp1 * cbet1;
+      calp0 = m.hypot(calp1, salp1 * sbet1); // calp0 > 0
+      if (calp0 !== 0 && salp0 !== 0) {
+        // From Lambda12: tan(bet) = tan(sig) * cos(alp)
+        ssig1 = sbet1; csig1 = calp1 * cbet1;
+        ssig2 = sbet2; csig2 = calp2 * cbet2;
+        k2 = m.sq(calp0) * this._ep2;
+        eps = k2 / (2 * (1 + Math.sqrt(1 + k2)) + k2);
+        // Multiplier = a^2 * e^2 * cos(alpha0) * sin(alpha0).
+        A4 = m.sq(this.a) * calp0 * salp0 * this._e2;
+        // norm(ssig1, csig1);
+        t = m.hypot(ssig1, csig1); ssig1 /= t; csig1 /= t;
+        // norm(ssig2, csig2);
+        t = m.hypot(ssig2, csig2); ssig2 /= t; csig2 /= t;
+        C4a = new Array(g.nC4_);
+        this.C4f(eps, C4a);
+        B41 = g.SinCosSeries(false, ssig1, csig1, C4a);
+        B42 = g.SinCosSeries(false, ssig2, csig2, C4a);
+        vals.S12 = A4 * (sig12 < 0.01 ?
+          g.SinCosSeries(false, ssig1, csig1, C4a, sig12) : B42 - B41);
+      } else
+        // Avoid problems with indeterminate sig1, sig2 on equator
+        vals.S12 = 0;
+      if (!meridian && somg12 == 2) {
+        somg12 = Math.sin(omg12); comg12 = Math.cos(omg12);
+      }
+      if (!meridian &&
+          comg12 > -0.7071 &&      // Long difference not too big
+          sbet2 - sbet1 < 1.75) { // Lat difference not too big
+        // Use tan(Gamma/2) = tan(omg12/2)
+        // * (tan(bet1/2)+tan(bet2/2))/(1+tan(bet1/2)*tan(bet2/2))
+        // with tan(x/2) = sin(x)/(1+cos(x))
+        domg12 = 1 + comg12; dbet1 = 1 + cbet1; dbet2 = 1 + cbet2;
+        alp12 = 2 * Math.atan2( somg12 * (sbet1*dbet2 + sbet2*dbet1),
+                                domg12 * (sbet1*sbet2 + dbet1*dbet2) );
+      } else {
+        // alp12 = alp2 - alp1, used in atan2 so no need to normalize
+        salp12 = salp2 * calp1 - calp2 * salp1;
+        calp12 = calp2 * calp1 + salp2 * salp1;
+        // The right thing appears to happen if alp1 = +/-180 and alp2 = 0, viz
+        // salp12 = -0 and alp12 = -180.  However this depends on the sign
+        // being attached to 0 correctly.  The following ensures the correct
+        // behavior.
+        if (salp12 === 0 && calp12 < 0) {
+          salp12 = g.tiny_ * calp1;
+          calp12 = -1;
+        }
+        alp12 = Math.atan2(salp12, calp12);
+      }
+      vals.S12 += this._c2 * alp12;
+      vals.S12 *= swapp * lonsign * latsign;
+      // Convert -0 to 0
+      vals.S12 += 0;
+    }
+
+    // Convert calp, salp to azimuth accounting for lonsign, swapp, latsign.
+    if (swapp < 0) {
+      [salp2, salp1] = [salp1, salp2]; // swap(salp1, salp2);
+      [calp2, calp1] = [calp1, calp2]; // swap(calp1, calp2);
+      if (outmask & g.GEODESICSCALE) {
+        [vals.M21, vals.M12] = [vals.M12, vals.M21]; //swap(vals.M12, vals.M21);
+      }
+    }
+
+    salp1 *= swapp * lonsign; calp1 *= swapp * latsign;
+    salp2 *= swapp * lonsign; calp2 *= swapp * latsign;
+
+    return {vals: vals,
+            salp1: salp1, calp1: calp1,
+            salp2: salp2, calp2: calp2};
+  };
+
+  /**
+   * @summary Solve the general direct geodesic problem.
+   * @param {number} lat1 the latitude of the first point in degrees.
+   * @param {number} lon1 the longitude of the first point in degrees.
+   * @param {number} azi1 the azimuth at the first point in degrees.
+   * @param {bool} arcmode is the next parameter an arc length?
+   * @param {number} s12_a12 the (arcmode ? arc length : distance) from the
+   *   first point to the second in (arcmode ? degrees : meters).
+   * @param {bitmask} [outmask = STANDARD] which results to include.
+   * @return {object} the requested results.
+   * @description The lat1, lon1, azi1, and a12 fields of the result are always
+   *   set; s12 is included if arcmode is false.  For details on the outmask
+   *   parameter, see {@tutorial 2-interface}, "The outmask and caps
+   *   parameters".
+   */
+  g.Geodesic.prototype.GenDirect = function(lat1, lon1, azi1,
+                                            arcmode, s12_a12, outmask) {
+    var line;
+    if (!outmask) outmask = g.STANDARD;
+    else if (outmask === g.LONG_UNROLL) outmask |= g.STANDARD;
+    // Automatically supply DISTANCE_IN if necessary
+    if (!arcmode) outmask |= g.DISTANCE_IN;
+    line = new l.GeodesicLine(this, lat1, lon1, azi1, outmask);
+    return line.GenPosition(arcmode, s12_a12, outmask);
+  };
+
+  /**
+   * @summary Solve the direct geodesic problem.
+   * @param {number} lat1 the latitude of the first point in degrees.
+   * @param {number} lon1 the longitude of the first point in degrees.
+   * @param {number} azi1 the azimuth at the first point in degrees.
+   * @param {number} s12 the distance from the first point to the second in
+   *   meters.
+   * @param {bitmask} [outmask = STANDARD] which results to include.
+   * @return {object} the requested results.
+   * @description The lat1, lon1, azi1, s12, and a12 fields of the result are
+   *   always set.  For details on the outmask parameter, see {@tutorial
+   *   2-interface}, "The outmask and caps parameters".
+   */
+  g.Geodesic.prototype.Direct = function(lat1, lon1, azi1, s12, outmask) {
+    return this.GenDirect(lat1, lon1, azi1, false, s12, outmask);
+  };
+
+  /**
+   * @summary Solve the direct geodesic problem with arc length.
+   * @param {number} lat1 the latitude of the first point in degrees.
+   * @param {number} lon1 the longitude of the first point in degrees.
+   * @param {number} azi1 the azimuth at the first point in degrees.
+   * @param {number} a12 the arc length from the first point to the second in
+   *   degrees.
+   * @param {bitmask} [outmask = STANDARD] which results to include.
+   * @return {object} the requested results.
+   * @description The lat1, lon1, azi1, and a12 fields of the result are
+   *   always set.  For details on the outmask parameter, see {@tutorial
+   *   2-interface}, "The outmask and caps parameters".
+   */
+  g.Geodesic.prototype.ArcDirect = function(lat1, lon1, azi1, a12, outmask) {
+    return this.GenDirect(lat1, lon1, azi1, true, a12, outmask);
+  };
+
+  /**
+   * @summary Create a {@link module:geodesic/GeodesicLine.GeodesicLine
+   *   GeodesicLine} object.
+   * @param {number} lat1 the latitude of the first point in degrees.
+   * @param {number} lon1 the longitude of the first point in degrees.
+   * @param {number} azi1 the azimuth at the first point in degrees.
+   *   degrees.
+   * @param {bitmask} [caps = STANDARD | DISTANCE_IN] which capabilities to
+   *   include.
+   * @return {object} the
+   *   {@link module:geodesic/GeodesicLine.GeodesicLine
+   *   GeodesicLine} object
+   * @description For details on the caps parameter, see {@tutorial
+   *   2-interface}, "The outmask and caps parameters".
+   */
+  g.Geodesic.prototype.Line = function(lat1, lon1, azi1, caps) {
+    return new l.GeodesicLine(this, lat1, lon1, azi1, caps);
+  };
+
+  /**
+   * @summary Define a {@link module:geodesic/GeodesicLine.GeodesicLine
+   *   GeodesicLine} in terms of the direct geodesic problem specified in terms
+   *   of distance.
+   * @param {number} lat1 the latitude of the first point in degrees.
+   * @param {number} lon1 the longitude of the first point in degrees.
+   * @param {number} azi1 the azimuth at the first point in degrees.
+   *   degrees.
+   * @param {number} s12 the distance between point 1 and point 2 (meters); it
+   *   can be negative.
+   * @param {bitmask} [caps = STANDARD | DISTANCE_IN] which capabilities to
+   *   include.
+   * @return {object} the
+   *   {@link module:geodesic/GeodesicLine.GeodesicLine
+   *   GeodesicLine} object
+   * @description This function sets point 3 of the GeodesicLine to correspond
+   *   to point 2 of the direct geodesic problem.  For details on the caps
+   *   parameter, see {@tutorial 2-interface}, "The outmask and caps
+   *   parameters".
+   */
+  g.Geodesic.prototype.DirectLine = function(lat1, lon1, azi1, s12, caps) {
+    return this.GenDirectLine(lat1, lon1, azi1, false, s12, caps);
+  };
+
+  /**
+   * @summary Define a {@link module:geodesic/GeodesicLine.GeodesicLine
+   *   GeodesicLine} in terms of the direct geodesic problem specified in terms
+   *   of arc length.
+   * @param {number} lat1 the latitude of the first point in degrees.
+   * @param {number} lon1 the longitude of the first point in degrees.
+   * @param {number} azi1 the azimuth at the first point in degrees.
+   *   degrees.
+   * @param {number} a12 the arc length between point 1 and point 2 (degrees);
+   *   it can be negative.
+   * @param {bitmask} [caps = STANDARD | DISTANCE_IN] which capabilities to
+   *   include.
+   * @return {object} the
+   *   {@link module:geodesic/GeodesicLine.GeodesicLine
+   *   GeodesicLine} object
+   * @description This function sets point 3 of the GeodesicLine to correspond
+   *   to point 2 of the direct geodesic problem.  For details on the caps
+   *   parameter, see {@tutorial 2-interface}, "The outmask and caps
+   *   parameters".
+   */
+  g.Geodesic.prototype.ArcDirectLine = function(lat1, lon1, azi1, a12, caps) {
+    return this.GenDirectLine(lat1, lon1, azi1, true, a12, caps);
+  };
+
+  /**
+   * @summary Define a {@link module:geodesic/GeodesicLine.GeodesicLine
+   *   GeodesicLine} in terms of the direct geodesic problem specified in terms
+   *   of either distance or arc length.
+   * @param {number} lat1 the latitude of the first point in degrees.
+   * @param {number} lon1 the longitude of the first point in degrees.
+   * @param {number} azi1 the azimuth at the first point in degrees.
+   *   degrees.
+   * @param {bool} arcmode boolean flag determining the meaning of the
+   *   s12_a12.
+   * @param {number} s12_a12 if arcmode is false, this is the distance between
+   *   point 1 and point 2 (meters); otherwise it is the arc length between
+   *   point 1 and point 2 (degrees); it can be negative.
+   * @param {bitmask} [caps = STANDARD | DISTANCE_IN] which capabilities to
+   *   include.
+   * @return {object} the
+   *   {@link module:geodesic/GeodesicLine.GeodesicLine
+   *   GeodesicLine} object
+   * @description This function sets point 3 of the GeodesicLine to correspond
+   *   to point 2 of the direct geodesic problem.  For details on the caps
+   *   parameter, see {@tutorial 2-interface}, "The outmask and caps
+   *   parameters".
+   */
+  g.Geodesic.prototype.GenDirectLine = function(lat1, lon1, azi1,
+                                                arcmode, s12_a12, caps) {
+    var t;
+    if (!caps) caps = g.STANDARD | g.DISTANCE_IN;
+    // Automatically supply DISTANCE_IN if necessary
+    if (!arcmode) caps |= g.DISTANCE_IN;
+    t = new l.GeodesicLine(this, lat1, lon1, azi1, caps);
+    t.GenSetDistance(arcmode, s12_a12);
+    return t;
+  };
+
+  /**
+   * @summary Define a {@link module:geodesic/GeodesicLine.GeodesicLine
+   *   GeodesicLine} in terms of the inverse geodesic problem.
+   * @param {number} lat1 the latitude of the first point in degrees.
+   * @param {number} lon1 the longitude of the first point in degrees.
+   * @param {number} lat2 the latitude of the second point in degrees.
+   * @param {number} lon2 the longitude of the second point in degrees.
+   * @param {bitmask} [caps = STANDARD | DISTANCE_IN] which capabilities to
+   *   include.
+   * @return {object} the
+   *   {@link module:geodesic/GeodesicLine.GeodesicLine
+   *   GeodesicLine} object
+   * @description This function sets point 3 of the GeodesicLine to correspond
+   *   to point 2 of the inverse geodesic problem.  For details on the caps
+   *   parameter, see {@tutorial 2-interface}, "The outmask and caps
+   *   parameters".
+   */
+  g.Geodesic.prototype.InverseLine = function(lat1, lon1, lat2, lon2, caps) {
+    var r, t, azi1;
+    if (!caps) caps = g.STANDARD | g.DISTANCE_IN;
+    r = this.InverseInt(lat1, lon1, lat2, lon2, g.ARC);
+    azi1 = m.atan2d(r.salp1, r.calp1);
+    // Ensure that a12 can be converted to a distance
+    if (caps & (g.OUT_MASK & g.DISTANCE_IN)) caps |= g.DISTANCE;
+    t = new l.GeodesicLine(this, lat1, lon1, azi1, caps, r.salp1, r.calp1);
+    t.SetArc(r.vals.a12);
+    return t;
+  };
+
+  /**
+   * @summary Create a {@link module:geodesic/PolygonArea.PolygonArea
+   *   PolygonArea} object.
+   * @param {bool} [polyline = false] if true the new PolygonArea object
+   *   describes a polyline instead of a polygon.
+   * @return {object} the
+   *   {@link module:geodesic/PolygonArea.PolygonArea
+   *   PolygonArea} object
+   */
+  g.Geodesic.prototype.Polygon = function(polyline) {
+    return new p.PolygonArea(this, polyline);
+  };
+
+  /**
+   * @summary a {@link module:geodesic/Geodesic.Geodesic Geodesic} object
+   *   initialized for the WGS84 ellipsoid.
+   * @constant {object}
+   */
+  g.WGS84 = new g.Geodesic(c.WGS84.a, c.WGS84.f);
+})(geodesic.Geodesic, geodesic.GeodesicLine,
+   geodesic.PolygonArea, geodesic.Math, geodesic.Constants);
+
+/*
+ * GeodesicLine.js
+ * Transcription of GeodesicLine.[ch]pp into JavaScript.
+ *
+ * See the documentation for the C++ class.  The conversion is a literal
+ * conversion from C++.
+ *
+ * The algorithms are derived in
+ *
+ *    Charles F. F. Karney,
+ *    Algorithms for geodesics, J. Geodesy 87, 43-55 (2013);
+ *    https://doi.org/10.1007/s00190-012-0578-z
+ *    Addenda: https://geographiclib.sourceforge.io/geod-addenda.html
+ *
+ * Copyright (c) Charles Karney (2011-2022) <karney@alum.mit.edu> and licensed
+ * under the MIT/X11 License.  For more information, see
+ * https://geographiclib.sourceforge.io/
+ */
+
+// Load AFTER geodesic/Math.js, geodesic/Geodesic.js
+
+(function(
+  g,
+  /**
+   * @exports geodesic/GeodesicLine
+   * @description Solve geodesic problems on a single geodesic line via the
+   *   {@link module:geodesic/GeodesicLine.GeodesicLine GeodesicLine}
+   *   class.
+   */
+  l, m) {
+  "use strict";
+
+  /**
+   * @class
+   * @property {number} a the equatorial radius (meters).
+   * @property {number} f the flattening.
+   * @property {number} lat1 the initial latitude (degrees).
+   * @property {number} lon1 the initial longitude (degrees).
+   * @property {number} azi1 the initial azimuth (degrees).
+   * @property {number} salp1 the sine of the azimuth at the first point.
+   * @property {number} calp1 the cosine the azimuth at the first point.
+   * @property {number} s13 the distance to point 3 (meters).
+   * @property {number} a13 the arc length to point 3 (degrees).
+   * @property {bitmask} caps the capabilities of the object.
+   * @summary Initialize a GeodesicLine object.  For details on the caps
+   *   parameter, see {@tutorial 2-interface}, "The outmask and caps
+   *   parameters".
+   * @classdesc Performs geodesic calculations along a given geodesic line.
+   *   This object is usually instantiated by
+   *   {@link module:geodesic/Geodesic.Geodesic#Line Geodesic.Line}.
+   *   The methods
+   *   {@link module:geodesic/Geodesic.Geodesic#DirectLine
+   *   Geodesic.DirectLine} and
+   *   {@link module:geodesic/Geodesic.Geodesic#InverseLine
+   *   Geodesic.InverseLine} set in addition the position of a reference point
+   *   3.
+   * @param {object} geod a {@link module:geodesic/Geodesic.Geodesic
+   *   Geodesic} object.
+   * @param {number} lat1 the latitude of the first point in degrees.
+   * @param {number} lon1 the longitude of the first point in degrees.
+   * @param {number} azi1 the azimuth at the first point in degrees.
+   * @param {bitmask} [caps = STANDARD | DISTANCE_IN] which capabilities to
+   *   include; LATITUDE | AZIMUTH are always included.
+   */
+  l.GeodesicLine = function(geod, lat1, lon1, azi1, caps, salp1, calp1) {
+    var t, cbet1, sbet1, eps, s, c;
+    if (!caps) caps = g.STANDARD | g.DISTANCE_IN;
+
+    this.a = geod.a;
+    this.f = geod.f;
+    this._b = geod._b;
+    this._c2 = geod._c2;
+    this._f1 = geod._f1;
+    this.caps = caps | g.LATITUDE | g.AZIMUTH | g.LONG_UNROLL;
+
+    this.lat1 = m.LatFix(lat1);
+    this.lon1 = lon1;
+    if (typeof salp1 === 'undefined' || typeof calp1 === 'undefined') {
+      this.azi1 = m.AngNormalize(azi1);
+      t = m.sincosd(m.AngRound(this.azi1)); this.salp1 = t.s; this.calp1 = t.c;
+    } else {
+      this.azi1 = azi1; this.salp1 = salp1; this.calp1 = calp1;
+    }
+    t = m.sincosd(m.AngRound(this.lat1)); sbet1 = this._f1 * t.s; cbet1 = t.c;
+    // norm(sbet1, cbet1);
+    t = m.hypot(sbet1, cbet1); sbet1 /= t; cbet1 /= t;
+    // Ensure cbet1 = +epsilon at poles
+    cbet1 = Math.max(g.tiny_, cbet1);
+    this._dn1 = Math.sqrt(1 + geod._ep2 * m.sq(sbet1));
+
+    // Evaluate alp0 from sin(alp1) * cos(bet1) = sin(alp0),
+    this._salp0 = this.salp1 * cbet1; // alp0 in [0, pi/2 - |bet1|]
+    // Alt: calp0 = hypot(sbet1, calp1 * cbet1).  The following
+    // is slightly better (consider the case salp1 = 0).
+    this._calp0 = m.hypot(this.calp1, this.salp1 * sbet1);
+    // Evaluate sig with tan(bet1) = tan(sig1) * cos(alp1).
+    // sig = 0 is nearest northward crossing of equator.
+    // With bet1 = 0, alp1 = pi/2, we have sig1 = 0 (equatorial line).
+    // With bet1 =  pi/2, alp1 = -pi, sig1 =  pi/2
+    // With bet1 = -pi/2, alp1 =  0 , sig1 = -pi/2
+    // Evaluate omg1 with tan(omg1) = sin(alp0) * tan(sig1).
+    // With alp0 in (0, pi/2], quadrants for sig and omg coincide.
+    // No atan2(0,0) ambiguity at poles since cbet1 = +epsilon.
+    // With alp0 = 0, omg1 = 0 for alp1 = 0, omg1 = pi for alp1 = pi.
+    this._ssig1 = sbet1; this._somg1 = this._salp0 * sbet1;
+    this._csig1 = this._comg1 =
+      sbet1 !== 0 || this.calp1 !== 0 ? cbet1 * this.calp1 : 1;
+    // norm(this._ssig1, this._csig1); // sig1 in (-pi, pi]
+    t = m.hypot(this._ssig1, this._csig1);
+    this._ssig1 /= t; this._csig1 /= t;
+    // norm(this._somg1, this._comg1); -- don't need to normalize!
+
+    this._k2 = m.sq(this._calp0) * geod._ep2;
+    eps = this._k2 / (2 * (1 + Math.sqrt(1 + this._k2)) + this._k2);
+
+    if (this.caps & g.CAP_C1) {
+      this._A1m1 = g.A1m1f(eps);
+      this._C1a = new Array(g.nC1_ + 1);
+      g.C1f(eps, this._C1a);
+      this._B11 = g.SinCosSeries(true, this._ssig1, this._csig1, this._C1a);
+      s = Math.sin(this._B11); c = Math.cos(this._B11);
+      // tau1 = sig1 + B11
+      this._stau1 = this._ssig1 * c + this._csig1 * s;
+      this._ctau1 = this._csig1 * c - this._ssig1 * s;
+      // Not necessary because C1pa reverts C1a
+      //    _B11 = -SinCosSeries(true, _stau1, _ctau1, _C1pa);
+    }
+
+    if (this.caps & g.CAP_C1p) {
+      this._C1pa = new Array(g.nC1p_ + 1);
+      g.C1pf(eps, this._C1pa);
+    }
+
+    if (this.caps & g.CAP_C2) {
+      this._A2m1 = g.A2m1f(eps);
+      this._C2a = new Array(g.nC2_ + 1);
+      g.C2f(eps, this._C2a);
+      this._B21 = g.SinCosSeries(true, this._ssig1, this._csig1, this._C2a);
+    }
+
+    if (this.caps & g.CAP_C3) {
+      this._C3a = new Array(g.nC3_);
+      geod.C3f(eps, this._C3a);
+      this._A3c = -this.f * this._salp0 * geod.A3f(eps);
+      this._B31 = g.SinCosSeries(true, this._ssig1, this._csig1, this._C3a);
+    }
+
+    if (this.caps & g.CAP_C4) {
+      this._C4a = new Array(g.nC4_); // all the elements of _C4a are used
+      geod.C4f(eps, this._C4a);
+      // Multiplier = a^2 * e^2 * cos(alpha0) * sin(alpha0)
+      this._A4 = m.sq(this.a) * this._calp0 * this._salp0 * geod._e2;
+      this._B41 = g.SinCosSeries(false, this._ssig1, this._csig1, this._C4a);
+    }
+
+    this.a13 = this.s13 = NaN;
+  };
+
+  /**
+   * @summary Find the position on the line (general case).
+   * @param {bool} arcmode is the next parameter an arc length?
+   * @param {number} s12_a12 the (arcmode ? arc length : distance) from the
+   *   first point to the second in (arcmode ? degrees : meters).
+   * @param {bitmask} [outmask = STANDARD] which results to include; this is
+   *   subject to the capabilities of the object.
+   * @return {object} the requested results.
+   * @description The lat1, lon1, azi1, and a12 fields of the result are
+   *   always set; s12 is included if arcmode is false.  For details on the
+   *   outmask parameter, see {@tutorial 2-interface}, "The outmask and caps
+   *   parameters".
+   */
+  l.GeodesicLine.prototype.GenPosition = function(arcmode, s12_a12,
+                                                  outmask, vals = {}) {
+    // math.gl: optional caller-owned result; numerical equations are unchanged.
+    var
+        sig12, ssig12, csig12, B12, AB1, ssig2, csig2, tau12, s, c, serr,
+        omg12, lam12, lon12, E, sbet2, cbet2, somg2, comg2, salp2, calp2, dn2,
+        B22, AB2, J12, t, B42, salp12, calp12;
+    if (!outmask) outmask = g.STANDARD;
+    else if (outmask === g.LONG_UNROLL) outmask |= g.STANDARD;
+    outmask &= this.caps & g.OUT_MASK;
+    vals.lat1 = this.lat1; vals.azi1 = this.azi1;
+    vals.lon1 = outmask & g.LONG_UNROLL ?
+      this.lon1 : m.AngNormalize(this.lon1);
+    if (arcmode)
+      vals.a12 = s12_a12;
+    else
+      vals.s12 = s12_a12;
+    if (!( arcmode || (this.caps & g.DISTANCE_IN & g.OUT_MASK) )) {
+      // Uninitialized or impossible distance calculation requested
+      vals.a12 = NaN;
+      return vals;
+    }
+
+    // Avoid warning about uninitialized B12.
+    B12 = 0; AB1 = 0;
+    if (arcmode) {
+      // Interpret s12_a12 as spherical arc length
+      sig12 = s12_a12 * m.degree;
+      t = m.sincosd(s12_a12); ssig12 = t.s; csig12 = t.c;
+    } else {
+      // Interpret s12_a12 as distance
+      tau12 = s12_a12 / (this._b * (1 + this._A1m1));
+      s = Math.sin(tau12);
+      c = Math.cos(tau12);
+      // tau2 = tau1 + tau12
+      B12 = -g.SinCosSeries(true,
+                            this._stau1 * c + this._ctau1 * s,
+                            this._ctau1 * c - this._stau1 * s,
+                            this._C1pa);
+      sig12 = tau12 - (B12 - this._B11);
+      ssig12 = Math.sin(sig12); csig12 = Math.cos(sig12);
+      if (Math.abs(this.f) > 0.01) {
+        // Reverted distance series is inaccurate for |f| > 1/100, so correct
+        // sig12 with 1 Newton iteration.  The following table shows the
+        // approximate maximum error for a = WGS_a() and various f relative to
+        // GeodesicExact.
+        //     erri = the error in the inverse solution (nm)
+        //     errd = the error in the direct solution (series only) (nm)
+        //     errda = the error in the direct solution
+        //             (series + 1 Newton) (nm)
+        //
+        //       f     erri  errd errda
+        //     -1/5    12e6 1.2e9  69e6
+        //     -1/10  123e3  12e6 765e3
+        //     -1/20   1110 108e3  7155
+        //     -1/50  18.63 200.9 27.12
+        //     -1/100 18.63 23.78 23.37
+        //     -1/150 18.63 21.05 20.26
+        //      1/150 22.35 24.73 25.83
+        //      1/100 22.35 25.03 25.31
+        //      1/50  29.80 231.9 30.44
+        //      1/20   5376 146e3  10e3
+        //      1/10  829e3  22e6 1.5e6
+        //      1/5   157e6 3.8e9 280e6
+        ssig2 = this._ssig1 * csig12 + this._csig1 * ssig12;
+        csig2 = this._csig1 * csig12 - this._ssig1 * ssig12;
+        B12 = g.SinCosSeries(true, ssig2, csig2, this._C1a);
+        serr = (1 + this._A1m1) * (sig12 + (B12 - this._B11)) -
+          s12_a12 / this._b;
+        sig12 = sig12 - serr / Math.sqrt(1 + this._k2 * m.sq(ssig2));
+        ssig12 = Math.sin(sig12); csig12 = Math.cos(sig12);
+        // Update B12 below
+      }
+    }
+
+    // sig2 = sig1 + sig12
+    ssig2 = this._ssig1 * csig12 + this._csig1 * ssig12;
+    csig2 = this._csig1 * csig12 - this._ssig1 * ssig12;
+    dn2 = Math.sqrt(1 + this._k2 * m.sq(ssig2));
+    if (outmask & (g.DISTANCE | g.REDUCEDLENGTH | g.GEODESICSCALE)) {
+      if (arcmode || Math.abs(this.f) > 0.01)
+        B12 = g.SinCosSeries(true, ssig2, csig2, this._C1a);
+      AB1 = (1 + this._A1m1) * (B12 - this._B11);
+    }
+    // sin(bet2) = cos(alp0) * sin(sig2)
+    sbet2 = this._calp0 * ssig2;
+    // Alt: cbet2 = hypot(csig2, salp0 * ssig2);
+    cbet2 = m.hypot(this._salp0, this._calp0 * csig2);
+    if (cbet2 === 0)
+      // I.e., salp0 = 0, csig2 = 0.  Break the degeneracy in this case
+      cbet2 = csig2 = g.tiny_;
+    // tan(alp0) = cos(sig2)*tan(alp2)
+    salp2 = this._salp0; calp2 = this._calp0 * csig2; // No need to normalize
+
+    if (arcmode && (outmask & g.DISTANCE))
+      vals.s12 = this._b * ((1 + this._A1m1) * sig12 + AB1);
+
+    if (outmask & g.LONGITUDE) {
+      // tan(omg2) = sin(alp0) * tan(sig2)
+      somg2 = this._salp0 * ssig2; comg2 = csig2; // No need to normalize
+      E = m.copysign(1, this._salp0);
+      // omg12 = omg2 - omg1
+      omg12 = outmask & g.LONG_UNROLL ?
+        E * (sig12 -
+             (Math.atan2(ssig2, csig2) -
+              Math.atan2(this._ssig1, this._csig1)) +
+             (Math.atan2(E * somg2, comg2) -
+              Math.atan2(E * this._somg1, this._comg1))) :
+        Math.atan2(somg2 * this._comg1 - comg2 * this._somg1,
+                     comg2 * this._comg1 + somg2 * this._somg1);
+      lam12 = omg12 + this._A3c *
+        ( sig12 + (g.SinCosSeries(true, ssig2, csig2, this._C3a) -
+                   this._B31));
+      lon12 = lam12 / m.degree;
+      vals.lon2 = outmask & g.LONG_UNROLL ? this.lon1 + lon12 :
+        m.AngNormalize(m.AngNormalize(this.lon1) + m.AngNormalize(lon12));
+    }
+
+    if (outmask & g.LATITUDE)
+      vals.lat2 = m.atan2d(sbet2, this._f1 * cbet2);
+
+    if (outmask & g.AZIMUTH)
+      vals.azi2 = m.atan2d(salp2, calp2);
+
+    if (outmask & (g.REDUCEDLENGTH | g.GEODESICSCALE)) {
+      B22 = g.SinCosSeries(true, ssig2, csig2, this._C2a);
+      AB2 = (1 + this._A2m1) * (B22 - this._B21);
+      J12 = (this._A1m1 - this._A2m1) * sig12 + (AB1 - AB2);
+      if (outmask & g.REDUCEDLENGTH)
+        // Add parens around (_csig1 * ssig2) and (_ssig1 * csig2) to ensure
+        // accurate cancellation in the case of coincident points.
+        vals.m12 = this._b * ((      dn2 * (this._csig1 * ssig2) -
+                               this._dn1 * (this._ssig1 * csig2)) -
+                              this._csig1 * csig2 * J12);
+      if (outmask & g.GEODESICSCALE) {
+        t = this._k2 * (ssig2 - this._ssig1) * (ssig2 + this._ssig1) /
+          (this._dn1 + dn2);
+        vals.M12 = csig12 +
+          (t * ssig2 - csig2 * J12) * this._ssig1 / this._dn1;
+        vals.M21 = csig12 -
+          (t * this._ssig1 - this._csig1 * J12) * ssig2 / dn2;
+      }
+    }
+
+    if (outmask & g.AREA) {
+      B42 = g.SinCosSeries(false, ssig2, csig2, this._C4a);
+      if (this._calp0 === 0 || this._salp0 === 0) {
+        // alp12 = alp2 - alp1, used in atan2 so no need to normalize
+        salp12 = salp2 * this.calp1 - calp2 * this.salp1;
+        calp12 = calp2 * this.calp1 + salp2 * this.salp1;
+      } else {
+        // tan(alp) = tan(alp0) * sec(sig)
+        // tan(alp2-alp1) = (tan(alp2) -tan(alp1)) / (tan(alp2)*tan(alp1)+1)
+        // = calp0 * salp0 * (csig1-csig2) / (salp0^2 + calp0^2 * csig1*csig2)
+        // If csig12 > 0, write
+        //   csig1 - csig2 = ssig12 * (csig1 * ssig12 / (1 + csig12) + ssig1)
+        // else
+        //   csig1 - csig2 = csig1 * (1 - csig12) + ssig12 * ssig1
+        // No need to normalize
+        salp12 = this._calp0 * this._salp0 *
+          (csig12 <= 0 ? this._csig1 * (1 - csig12) + ssig12 * this._ssig1 :
+           ssig12 * (this._csig1 * ssig12 / (1 + csig12) + this._ssig1));
+        calp12 = m.sq(this._salp0) + m.sq(this._calp0) * this._csig1 * csig2;
+      }
+      vals.S12 = this._c2 * Math.atan2(salp12, calp12) +
+        this._A4 * (Math.abs(sig12) < 0.01 ?
+          g.SinCosSeries(false, this._ssig1, this._csig1, this._C4a, sig12) : B42 - this._B41);
+    }
+
+    if (!arcmode)
+      vals.a12 = sig12 / m.degree;
+    return vals;
+  };
+
+  /**
+   * @summary Find the position on the line given s12.
+   * @param {number} s12 the distance from the first point to the second in
+   *   meters.
+   * @param {bitmask} [outmask = STANDARD] which results to include; this is
+   *   subject to the capabilities of the object.
+   * @return {object} the requested results.
+   * @description The lat1, lon1, azi1, s12, and a12 fields of the result are
+   *   always set; s12 is included if arcmode is false.  For details on the
+   *   outmask parameter, see {@tutorial 2-interface}, "The outmask and caps
+   *   parameters".
+   */
+  l.GeodesicLine.prototype.Position = function(s12, outmask) {
+    return this.GenPosition(false, s12, outmask);
+  };
+
+  /**
+   * @summary Find the position on the line given a12.
+   * @param {number} a12 the arc length from the first point to the second in
+   *   degrees.
+   * @param {bitmask} [outmask = STANDARD] which results to include; this is
+   *   subject to the capabilities of the object.
+   * @return {object} the requested results.
+   * @description The lat1, lon1, azi1, and a12 fields of the result are
+   *   always set.  For details on the outmask parameter, see {@tutorial
+   *   2-interface}, "The outmask and caps parameters".
+   */
+  l.GeodesicLine.prototype.ArcPosition = function(a12, outmask) {
+    return this.GenPosition(true, a12, outmask);
+  };
+
+  /**
+   * @summary Specify position of point 3 in terms of either distance or arc
+   *   length.
+   * @param {bool} arcmode boolean flag determining the meaning of the second
+   *   parameter; if arcmode is false, then the GeodesicLine object must have
+   *   been constructed with caps |= DISTANCE_IN.
+   * @param {number} s13_a13 if arcmode is false, this is the distance from
+   *   point 1 to point 3 (meters); otherwise it is the arc length from
+   *   point 1 to point 3 (degrees); it can be negative.
+   */
+  l.GeodesicLine.prototype.GenSetDistance = function(arcmode, s13_a13) {
+    if (arcmode)
+      this.SetArc(s13_a13);
+    else
+      this.SetDistance(s13_a13);
+  };
+
+  /**
+   * @summary Specify position of point 3 in terms distance.
+   * @param {number} s13 the distance from point 1 to point 3 (meters); it
+   *   can be negative.
+   */
+  l.GeodesicLine.prototype.SetDistance = function(s13) {
+    var r;
+    this.s13 = s13;
+    r = this.GenPosition(false, this.s13, g.ARC);
+    this.a13 = 0 + r.a12;       // the 0+ converts undefined into NaN
+  };
+
+  /**
+   * @summary Specify position of point 3 in terms of arc length.
+   * @param {number} a13 the arc length from point 1 to point 3 (degrees);
+   *   it can be negative.
+   */
+  l.GeodesicLine.prototype.SetArc = function(a13) {
+    var r;
+    this.a13 = a13;
+    r = this.GenPosition(true, this.a13, g.DISTANCE);
+    this.s13 = 0 + r.s12;       // the 0+ converts undefined into NaN
+  };
+
+})(geodesic.Geodesic, geodesic.GeodesicLine, geodesic.Math);
+
+export {geodesic};