Skip to content

supaernova.analysis.dispersion

[docs] module supaernova.analysis.dispersion

  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
from typing import TYPE_CHECKING, Literal
from pathlib import Path

import numpy as np

from .spectra import SpectraPlot, SpectraPlotter
from .analysis import Plotter

if TYPE_CHECKING:
    import pandas as pd

    from supaernova.configs.steps.data import DataStepResult
    from supaernova.configs.steps.posterior import PosteriorStepResult

    from .analysis import Axis, Figure


class DispersionPlot(SpectraPlot):
    subset: Literal["train", "test"]
    legacy: tuple[Path, ...] | None = None
    twins: str | None = None


class DispersionPlotter(Plotter):
    @staticmethod
    def plot_dispersion(
        data: "DataStepResult",
        hmcs: "list[PosteriorStepResult]",
        config: "DispersionPlot",
        *,
        fig: "Figure | None" = None,
        ax: "Axis | None" = None,
        twins: "pd.DataFrame | None" = None,
        legacy: "dict[str, np.ndarray] | None" = None,
        force: bool = False,
    ) -> None:
        savepath = (config.savepath or Path()) / f"{config.name}.{config.ext}"
        if savepath.exists() and not force:
            return

        _wl, _amplitude, _sigma, sn_mask, _spec_mask, _wl_mask = SpectraPlotter.prep(
            data, config
        )

        sn_mask = sn_mask[:, 0, 0]
        names = data.sn_name[:, 0, 0]

        twins_mask = np.ones_like(sn_mask)
        if twins is not None:
            twins_mask = np.zeros_like(sn_mask)
            twins_names = twins.name
            intersection = set(names) & set(twins_names)
            for name in intersection:
                ind = np.argwhere(names == name)[0]
                df = twins[twins.name == name]
                twins_mask[ind] = df.mask_twins

        redshift = data.redshift[:, 0, 0]
        order = np.argsort(redshift)
        redshift = redshift[order]
        redshift_error = (redshift * 3e5 + 300.0) / 3e5
        magshift_error = abs(-5 * np.log10(redshift / redshift_error))

        amplitudes = np.concatenate(
            [np.mean(hmc.hmc.delta_m, axis=0, keepdims=True) for hmc in hmcs], axis=0
        )[..., 0]
        amplitude_stds = np.concatenate(
            [
                np.sqrt(np.square(np.std(hmc.hmc.delta_m, axis=0, keepdims=True)))
                for hmc in hmcs
            ],
            axis=0,
        )[..., 0]

        weights = 1 / (amplitude_stds * amplitude_stds)
        weights /= weights.sum(axis=0)

        weighted_amplitudes = np.sum(weights * amplitudes, axis=0)
        weighted_amplitudes = weighted_amplitudes[order]

        weighted_variances = np.sum(
            weights * np.square(amplitudes - np.mean(amplitudes)), axis=0
        )
        weighted_variances = weighted_variances[order]

        weighted_stds = np.sqrt(
            (weighted_variances * weighted_variances)
            + (magshift_error * magshift_error)
        )

        def _plot(x, y, yerr, fig, ax, color, marker, alpha, title):
            k = 1.4826

            w_rms = np.sqrt(np.sum(y * y) / np.size(y))
            w_mad = np.std(y) / k

            print(title, x.shape, y.shape, yerr.shape)
            print("RMS: ", w_rms)
            print("NMAD: ", w_mad)

            fig, ax = Plotter.errorbar(
                x,
                y,
                yerr=yerr,
                fig=fig,
                ax=ax,
                color=color,
                marker=marker,
                alpha=alpha,
                label=f"{title} ({np.size(y)} SN) - RMS: {w_rms:.2f}, MAD: {w_mad:.2f}",
            )

            return fig, ax

        # === No Mask ===
        x = redshift
        y = weighted_amplitudes
        yerr = weighted_stds
        fig, ax = _plot(x, y, yerr, fig, ax, "black", "o", 0.25, "No Mask")

        # === SN Mask ===
        mask = sn_mask[order].astype(bool)
        x = redshift[mask]
        y = weighted_amplitudes[mask]
        yerr = weighted_stds[mask]
        fig, ax = _plot(x, y, yerr, fig, ax, "brown", "o", 0.25, "SN Mask")

        # === Twins Mask ===
        mask = twins_mask[order].astype(bool)
        x = redshift[mask]
        y = weighted_amplitudes[mask]
        yerr = weighted_stds[mask]
        fig, ax = _plot(x, y, yerr, fig, ax, "blue", "o", 0.25, "Twins Mask")

        # === Combined Mask ===
        mask = (sn_mask[order] * twins_mask[order]).astype(bool)
        x = redshift[mask]
        y = weighted_amplitudes[mask]
        yerr = weighted_stds[mask]
        fig, ax = _plot(x, y, yerr, fig, ax, "green", "o", 1, "Final")

        if legacy is not None:
            legacy_names = legacy["names"]
            intersection = set(names) & set(legacy_names)
            legacy_mask = np.zeros(legacy_names.shape, dtype=np.int32)
            for name in intersection:
                ind = np.argwhere(legacy_names == name)[0]
                legacy_mask[ind] = 1
            legacy_mask = legacy_mask.astype(bool)

            redshift = legacy["redshift"][legacy_mask]
            order = np.argsort(redshift)
            redshift = redshift[order]
            redshift_error = (redshift * 3e5 + 300.0) / 3e5
            magshift_error = abs(-5 * np.log10(redshift / redshift_error))

            amplitudes = legacy["amplitude_mcmc"][legacy_mask][None, ...]
            amplitude_stds = legacy["amplitude_mcmc_err"][legacy_mask][None, ...]

            weights = 1 / (amplitude_stds * amplitude_stds)
            weights /= weights.sum(axis=0)

            weighted_amplitudes = np.sum(weights * amplitudes, axis=0)
            weighted_amplitudes = amplitudes[0, ...][order]

            weighted_variances = np.sum(
                weights * np.square(amplitudes - np.mean(amplitudes)), axis=0
            )
            weighted_variances = weighted_variances[order]

            weighted_stds = np.sqrt(
                (weighted_variances * weighted_variances)
                + (magshift_error * magshift_error)
            )

            redshift_mask = ((redshift > 0.02) & (redshift < 1.0)).astype(int)
            sn_mask = redshift_mask * np.max(legacy["mask"][legacy_mask], axis=(-2, -1))

            twins_mask = np.ones_like(sn_mask)
            if twins is not None:
                twins_mask = np.zeros_like(sn_mask)
                names = legacy["names"][legacy_mask]
                twins_names = twins.name
                intersection = set(names) & set(twins_names)
                for name in intersection:
                    ind = np.argwhere(names == name)[0]
                    df = twins[twins.name == name]
                    twins_mask[ind] = df.mask_twins

            # === No Mask ===
            x = redshift
            y = weighted_amplitudes
            yerr = weighted_stds
            fig, ax = _plot(x, y, yerr, fig, ax, "black", "s", 0.25, "Legacy No Mask")

            # === SN Mask ===
            mask = sn_mask[order].astype(bool)
            x = redshift[mask]
            y = weighted_amplitudes[mask]
            yerr = weighted_stds[mask]
            fig, ax = _plot(x, y, yerr, fig, ax, "brown", "s", 0.25, "Legacy SN Mask")

            # === Twins Mask ===
            mask = twins_mask[order].astype(bool)
            x = redshift[mask]
            y = weighted_amplitudes[mask]
            yerr = weighted_stds[mask]
            fig, ax = _plot(x, y, yerr, fig, ax, "blue", "s", 0.25, "Legacy Twins Mask")

            # === Combined Mask ===
            mask = (sn_mask[order] * twins_mask[order]).astype(bool)
            x = redshift[mask]
            y = weighted_amplitudes[mask]
            yerr = weighted_stds[mask]
            fig, ax = _plot(x, y, yerr, fig, ax, "green", "s", 1, "Legacy Final")

        fig, ax = Plotter.axhline(0, fig=fig, ax=ax, color="black")
        ax.set_ylim(-0.75, 0.75)
        # ax.legend()
        fig = Plotter.save(fig, savepath)
        Plotter.close(fig, ax)