zodiacal 0.4.1

A blind astrometry plate-solving library
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313

use super::sphere;
use super::tan::TanWcs;

/// SIP (Simple Imaging Polynomial) WCS distortion model.
///
/// Extends the standard TAN projection with polynomial distortion terms
/// that model optical aberrations such as barrel/pincushion distortion.
///
/// Forward direction (pixel -> sky):
///   u, v = pixel offset from crpix
///   U = u + sum(A[p][q] * u^p * v^q)  for p+q >= 2
///   V = v + sum(B[p][q] * u^p * v^q)  for p+q >= 2
///   Then CD * (U, V) -> intermediate world coordinates -> sphere
///
/// Inverse direction (sky -> pixel):
///   Get (U, V) from TAN inverse projection
///   u = U + sum(AP[p][q] * U^p * V^q)  for p+q >= 2
///   v = V + sum(BP[p][q] * U^p * V^q)  for p+q >= 2
///   pixel = (u + crpix[0], v + crpix[1])
#[derive(Debug, Clone)]
pub struct SipWcs {
    /// Underlying TAN WCS.
    pub tan: TanWcs,
    /// Forward distortion coefficients for U.
    pub a: Vec<Vec<f64>>,
    /// Forward distortion coefficients for V.
    pub b: Vec<Vec<f64>>,
    /// Order of forward A polynomial.
    pub a_order: usize,
    /// Order of forward B polynomial.
    pub b_order: usize,
    /// Inverse distortion coefficients for u.
    pub ap: Vec<Vec<f64>>,
    /// Inverse distortion coefficients for v.
    pub bp: Vec<Vec<f64>>,
    /// Order of inverse AP polynomial.
    pub ap_order: usize,
    /// Order of inverse BP polynomial.
    pub bp_order: usize,
}

/// Evaluate a SIP polynomial: sum(coeffs[p][q] * u^p * v^q) for p+q >= 2.
fn eval_polynomial(coeffs: &[Vec<f64>], order: usize, u: f64, v: f64) -> f64 {
    let mut result = 0.0;
    let mut u_pow = 1.0; // u^p
    for p in 0..=order {
        let mut v_pow = 1.0; // v^q
        for q in 0..=order {
            if p + q >= 2 && p < coeffs.len() && q < coeffs[p].len() {
                result += coeffs[p][q] * u_pow * v_pow;
            }
            v_pow *= v;
        }
        u_pow *= u;
    }
    result
}

/// Create a zero-filled 2D coefficient array of size (order+1) x (order+1).
pub fn zero_coeffs(order: usize) -> Vec<Vec<f64>> {
    vec![vec![0.0; order + 1]; order + 1]
}

impl SipWcs {
    /// Create an identity SIP WCS (zero distortion coefficients) from a TAN WCS.
    pub fn from_tan(tan: TanWcs, order: usize) -> Self {
        SipWcs {
            a: zero_coeffs(order),
            b: zero_coeffs(order),
            a_order: order,
            b_order: order,
            ap: zero_coeffs(order),
            bp: zero_coeffs(order),
            ap_order: order,
            bp_order: order,
            tan,
        }
    }

    /// Convert pixel coordinates to a unit vector on the celestial sphere.
    pub fn pixel_to_xyz(&self, px: f64, py: f64) -> [f64; 3] {
        let u = px - self.tan.crpix[0];
        let v = py - self.tan.crpix[1];

        // Apply forward SIP distortion.
        let u_dist = u + eval_polynomial(&self.a, self.a_order, u, v);
        let v_dist = v + eval_polynomial(&self.b, self.b_order, u, v);

        // Apply CD matrix to get intermediate world coordinates.
        let x = self.tan.cd[0][0] * u_dist + self.tan.cd[0][1] * v_dist;
        let y = self.tan.cd[1][0] * u_dist + self.tan.cd[1][1] * v_dist;

        self.tan.iwc_to_xyz(x, y)
    }

    /// Convert a unit vector on the celestial sphere to pixel coordinates.
    ///
    /// Returns `None` if the point is behind the tangent plane.
    pub fn xyz_to_pixel(&self, xyz: [f64; 3]) -> Option<(f64, f64)> {
        let reference = sphere::radec_to_xyz(self.tan.crval[0], self.tan.crval[1]);
        let (x, y) = sphere::star_coords(xyz, reference)?;

        // Invert CD matrix to get undistorted pixel offsets (U, V).
        let det = self.tan.cd[0][0] * self.tan.cd[1][1] - self.tan.cd[0][1] * self.tan.cd[1][0];
        let inv_det = 1.0 / det;
        let u_big = inv_det * (self.tan.cd[1][1] * x - self.tan.cd[0][1] * y);
        let v_big = inv_det * (-self.tan.cd[1][0] * x + self.tan.cd[0][0] * y);

        // Apply inverse SIP polynomial to recover original pixel offsets.
        let u = u_big + eval_polynomial(&self.ap, self.ap_order, u_big, v_big);
        let v = v_big + eval_polynomial(&self.bp, self.bp_order, u_big, v_big);

        Some((u + self.tan.crpix[0], v + self.tan.crpix[1]))
    }

    /// Convert pixel coordinates to (RA, Dec) in radians.
    pub fn pixel_to_radec(&self, px: f64, py: f64) -> (f64, f64) {
        sphere::xyz_to_radec(self.pixel_to_xyz(px, py))
    }

    /// Convert (RA, Dec) in radians to pixel coordinates.
    ///
    /// Returns `None` if the position is behind the tangent plane.
    pub fn radec_to_pixel(&self, ra: f64, dec: f64) -> Option<(f64, f64)> {
        self.xyz_to_pixel(sphere::radec_to_xyz(ra, dec))
    }

    /// Approximate pixel scale in degrees per pixel from the CD matrix determinant.
    pub fn pixel_scale(&self) -> f64 {
        self.tan.pixel_scale()
    }

    /// RA, Dec (radians) of the image center pixel.
    pub fn field_center(&self) -> (f64, f64) {
        self.pixel_to_radec(self.tan.image_size[0] / 2.0, self.tan.image_size[1] / 2.0)
    }

    /// Angular radius (radians) of the smallest circle centered on the image
    /// center that encloses all four corners.
    pub fn field_radius(&self) -> f64 {
        let (cx, cy) = (self.tan.image_size[0] / 2.0, self.tan.image_size[1] / 2.0);
        let center = self.pixel_to_xyz(cx, cy);
        let w = self.tan.image_size[0];
        let h = self.tan.image_size[1];

        [
            self.pixel_to_xyz(0.0, 0.0),
            self.pixel_to_xyz(w, 0.0),
            self.pixel_to_xyz(0.0, h),
            self.pixel_to_xyz(w, h),
        ]
        .iter()
        .map(|c| sphere::angular_distance(center, *c))
        .fold(0.0_f64, f64::max)
    }
}

#[cfg(test)]
mod tests {
    use super::*;
    use std::f64::consts::PI;

    const EPS: f64 = 1e-10;

    fn assert_close(a: f64, b: f64, tol: f64) {
        assert!(
            (a - b).abs() < tol,
            "expected {a} ~= {b} (diff = {})",
            (a - b).abs()
        );
    }

    fn test_tan_wcs() -> TanWcs {
        let arcsec_rad = (1.0_f64 / 3600.0).to_radians();
        TanWcs {
            crval: [PI, 0.25],
            crpix: [512.0, 512.0],
            cd: [[arcsec_rad, 0.0], [0.0, arcsec_rad]],
            image_size: [1024.0, 1024.0],
        }
    }

    #[test]
    fn identity_sip_matches_tan() {
        let tan = test_tan_wcs();
        let sip = SipWcs::from_tan(tan.clone(), 3);

        for &(px, py) in &[
            (512.0, 512.0),
            (0.0, 0.0),
            (1024.0, 1024.0),
            (256.0, 768.0),
            (100.0, 900.0),
        ] {
            let tan_xyz = tan.pixel_to_xyz(px, py);
            let sip_xyz = sip.pixel_to_xyz(px, py);
            for i in 0..3 {
                assert_close(tan_xyz[i], sip_xyz[i], EPS);
            }

            let (tan_ra, tan_dec) = tan.pixel_to_radec(px, py);
            let (sip_ra, sip_dec) = sip.pixel_to_radec(px, py);
            assert_close(tan_ra, sip_ra, EPS);
            assert_close(tan_dec, sip_dec, EPS);

            let (tan_px, tan_py) = tan.xyz_to_pixel(tan_xyz).unwrap();
            let (sip_px, sip_py) = sip.xyz_to_pixel(tan_xyz).unwrap();
            assert_close(tan_px, sip_px, 1e-6);
            assert_close(tan_py, sip_py, 1e-6);
        }
    }

    #[test]
    fn barrel_distortion_round_trip() {
        let tan = test_tan_wcs();
        let mut sip = SipWcs::from_tan(tan, 3);

        // Apply known barrel distortion: A[2][0] and B[0][2] coefficients.
        let k = 1e-6;
        sip.a[2][0] = k;
        sip.b[0][2] = k;

        // Compute matching inverse coefficients.
        // For small distortion, AP ~ -A and BP ~ -B to first order.
        sip.ap[2][0] = -k;
        sip.bp[0][2] = -k;

        // Test round-trip at various positions.
        for &(px, py) in &[
            (512.0, 512.0),
            (256.0, 256.0),
            (768.0, 768.0),
            (100.0, 512.0),
            (512.0, 100.0),
            (200.0, 800.0),
        ] {
            let (ra, dec) = sip.pixel_to_radec(px, py);
            let (px2, py2) = sip.radec_to_pixel(ra, dec).unwrap();
            assert_close(px, px2, 0.1);
            assert_close(py, py2, 0.1);
        }
    }

    #[test]
    fn pixel_scale_at_center_matches_tan() {
        let tan = test_tan_wcs();
        let sip = SipWcs::from_tan(tan.clone(), 3);

        assert_close(sip.pixel_scale(), tan.pixel_scale(), 1e-15);
    }

    #[test]
    fn distortion_moves_edge_pixels() {
        let tan = test_tan_wcs();
        let mut sip = SipWcs::from_tan(tan.clone(), 3);

        let k = 1e-6;
        sip.a[2][0] = k;
        sip.b[0][2] = k;

        // At crpix, distortion should be zero since u=v=0.
        let sip_center = sip.pixel_to_xyz(512.0, 512.0);
        let tan_center = tan.pixel_to_xyz(512.0, 512.0);
        for i in 0..3 {
            assert_close(sip_center[i], tan_center[i], EPS);
        }

        // At corner, distortion should shift the result.
        let sip_corner = sip.pixel_to_xyz(0.0, 0.0);
        let tan_corner = tan.pixel_to_xyz(0.0, 0.0);
        let dist = sphere::angular_distance(sip_corner, tan_corner);
        assert!(
            dist > 1e-8,
            "expected distortion to move corner, got dist = {dist}"
        );
    }

    #[test]
    fn eval_polynomial_zero_coeffs() {
        let coeffs = zero_coeffs(3);
        let val = eval_polynomial(&coeffs, 3, 100.0, 200.0);
        assert_close(val, 0.0, EPS);
    }

    #[test]
    fn eval_polynomial_known_value() {
        // Set A[2][0] = 1.0 => result should be u^2.
        let mut coeffs = zero_coeffs(3);
        coeffs[2][0] = 1.0;
        let val = eval_polynomial(&coeffs, 3, 5.0, 3.0);
        assert_close(val, 25.0, EPS);

        // Set A[1][1] = 2.0 additionally => result = u^2 + 2*u*v.
        coeffs[1][1] = 2.0;
        let val = eval_polynomial(&coeffs, 3, 5.0, 3.0);
        assert_close(val, 25.0 + 30.0, EPS);
    }

    #[test]
    fn field_center_and_radius() {
        let tan = test_tan_wcs();
        let sip = SipWcs::from_tan(tan.clone(), 3);

        let (sip_ra, sip_dec) = sip.field_center();
        let (tan_ra, tan_dec) = tan.field_center();
        assert_close(sip_ra, tan_ra, EPS);
        assert_close(sip_dec, tan_dec, EPS);

        let sip_r = sip.field_radius();
        let tan_r = tan.field_radius();
        assert_close(sip_r, tan_r, 1e-8);
    }
}