def plot_observations(ax,
key,
range,
bins=10,
marker="o",
color="k",
markersize=3,
linewidth=0):
""" TODO: Move to utils.
"""
df = load_sample(select=True)
values = df[key].values
h, e = np.histogram(values, range=range, bins=bins)
c = 0.5 * (e[1:] + e[:-1])
err = np.sqrt(h)
norm = h.sum() * (e[1] - e[0])
xlim = ax.get_xlim()
ylim = ax.get_ylim()
ax.errorbar(c,
h / norm,
yerr=err / norm,
elinewidth=1,
markersize=markersize,
color=color,
linewidth=linewidth,
marker=marker,
label="observed",
zorder=10)
ax.set_xlim(xlim)
ax.set_ylim(ylim)
return ax
def make_faux_observations(self,
df=None,
dfo=None,
makeplots=False,
**kwargs):
""" Get sample of mock observations from the model sample
Args:
df (pd.DataFrame): Model samples from the best-fitting model.
Returns:
pd.DataFrame: Faux observations extracted from model samples.
"""
if df is None:
df = self.load_best_sample(**kwargs)
if dfo is None:
dfo = load_sample(select=True)
# Choose a random sample of the same size as observations
indices = np.random.randint(0, df.shape[0], dfo.shape[0])
# Map the model keys into observation keys
dff = pd.DataFrame()
for key, obskey in self.config["obskeys"].items():
dff[obskey] = df[key].values[indices]
if makeplots:
self.summary_plot(dfo=dff)
return dff
def plot_observations(grid, ax_dict, color="k"):
"""
"""
dfo = load_sample(select=True)
# Find the good-fitting model with the best ks-statistics
dfm = grid.load_confident_metrics()
metrics = dfm.iloc[np.argmax(dfm["kstest_min"].values)]
for key, ax in ax_dict.items():
if key in OBSKEYS:
ax = ax_dict[key]
values = dfo[OBSKEYS[key]].values
rng = RANGES[key]
hist_norm, edges = np.histogram(values,
range=rng,
bins=BINS_OBS,
density=True)
centres = 0.5 * (edges[1:] + edges[:-1])
hist, _ = np.histogram(values,
range=rng,
bins=BINS_OBS,
density=False)
yerr = np.sqrt(hist) * hist_norm.max() / hist.max()
ax.errorbar(centres,
hist_norm,
yerr=yerr,
color=color,
linewidth=0,
linestyle=None,
elinewidth=1.5,
marker="o",
markersize=3,
zorder=10)
pval = metrics[KSTEST_KEYS[key]]
ax.text(0.22,
0.75,
r"$p_{KS}=$" + rf"{pval:.2f}",
transform=ax.transAxes,
fontsize=FONTSIZE - 1,
color="k")
def __init__(self, *args, **kwargs):
super().__init__(*args, **kwargs)
self._dfobs = load_sample(config=self.config,
logger=self.logger,
select=True)
return result
if __name__ == "__main__":
model_name = "blue_final"
metric = "poisson_likelihood_2d"
confs = None
config = get_config()
image_dir = os.path.join(config["directories"]["data"], "images")
image_filename = os.path.join(image_dir, f"model_fit_{model_name}.png")
grid = ParameterGrid(model_name)
dfm = grid.load_best_sample(metric=metric)
dfo = load_sample()
keys = "uae_obs_jig", "rec_obs_jig"
range = {k: r for k, r in zip(keys, RANGE)}
dfs = grid.load_confident_samples(q=Q)
dfbest = grid.load_best_sample()
confs = get_confidence_intervals(dfs, dfbest, keys, range=range, bins=BINS1D)
fig = plt.figure(figsize=FIGSIZE)
spec = gridspec.GridSpec(ncols=10, nrows=2, figure=fig, hspace=0.35, wspace=1.6)
ax1 = fig.add_subplot(spec[:, :4])
ax2 = fig.add_subplot(spec[0, 4:])
ax3 = fig.add_subplot(spec[1, 4:])
# 2D model histogram
import os
from udgsizes.core import get_config, get_logger
from udgsizes.obs.sample import load_sample, load_gama_specobj
from udgsizes.utils import xmatch
if __name__ == "__main__":
logger = get_logger()
radius = 3. / 3600
df = load_sample(select=False)
dfg = load_gama_specobj()
dfm = xmatch.match_dataframe(df, dfg, radius=radius)
logger.info(f"Matched {dfm.shape[0]} sources.")
datadir = get_config()["directories"]["data"]
dfm.to_csv(os.path.join(datadir, "input", "lsbgs_gama_xmatch.csv"))
def fit_summary_plot(df, dfo=None, show=True, bins=15, select=True, **kwargs):
if dfo is None:
dfo = load_sample(select=select, **kwargs)
if select:
df = df[df["selected_jig"] == 1].reset_index(drop=True)
fig = plt.figure(figsize=(6, 6))
ax0 = plt.subplot(3, 1, 1)
histkwargs = dict(density=True, histtype="step")
rng = (min(dfo['mueff_av'].min(), df['uae_obs_jig'].min()),
max(dfo['mueff_av'].max(), df['uae_obs_jig'].max()))
ax0.hist(dfo['mueff_av'].values,
color="k",
range=rng,
bins=bins,
label="obs",
**histkwargs)
ax0.hist(df['uae_obs_jig'].values,
color="b",
range=rng,
bins=bins,
label="model",
**histkwargs)
ks = kstest(dfo['mueff_av'].values, df['uae_obs_jig'].values)[1]
ax0.legend(loc="best")
ax0.set_xlabel(f"uae (KS pval={ks:.2f})")
ax1 = plt.subplot(3, 1, 2)
rng = (min(dfo['rec_arcsec'].min(), df['rec_obs_jig'].min()),
max(dfo['rec_arcsec'].max(), df['rec_obs_jig'].max()))
ax1.hist(dfo['rec_arcsec'].values,
color="k",
range=rng,
bins=bins,
**histkwargs,
label="obs")
ax1.hist(df['rec_obs_jig'].values,
color="b",
range=rng,
bins=bins,
**histkwargs,
label="model")
ks = kstest(dfo['rec_arcsec'].values, df['rec_obs_jig'].values)[1]
ax1.legend(loc="best")
ax1.set_xlabel(f"rec (KS pval={ks:.2f})")
ax1 = plt.subplot(3, 1, 3)
rng = (min(dfo['g_r'].min(), df['colour_obs'].min()),
max(dfo['g_r'].max(), df['colour_obs'].max()))
ax1.hist(dfo['g_r'].values,
color="k",
range=rng,
bins=bins,
**histkwargs,
label="obs")
ax1.hist(df['colour_obs'].values,
color="b",
range=rng,
bins=bins,
**histkwargs,
label="model")
ks = kstest(dfo['g_r'].values, df['colour_obs'].values)[1]
ax1.legend(loc="best")
ax1.set_xlabel(f"g-r (KS pval={ks:.2f})")
plt.tight_layout()
if show:
plt.show(block=False)
return fig
OBSKEYS = {
"uae_obs_jig": "mueff_av",
"rec_obs_jig": "rec_arcsec",
"colour_obs": "g_r"
}
# TODO: Add KS values to graphs
# TODO: Make into grid method / plotting utils
if __name__ == "__main__":
grid = ParameterGrid(MODEL_NAME)
df = grid.load_best_sample(apply_prior=True, select=True)
df = df[df["selected_jig"].values == 1].reset_index(drop=True)
dfo = load_sample(select=True)
fig = plt.figure(figsize=(FIGHEIGHT * len(PAR_NAMES), FIGHEIGHT * 1.2))
for i, par_name in enumerate(PAR_NAMES):
ax = plt.subplot(1, len(PAR_NAMES), i + 1)
values = df[par_name].values
ax.hist(values, color="k", alpha=0.4, **HISTKWARGS)
if par_name in OBSKEYS:
values_obs = dfo[OBSKEYS[par_name]].values
rng = values.min(), values.max()
ax.hist(values_obs, range=rng, color="b", alpha=0.4, **HISTKWARGS)
if i == 0:
import matplotlib.pyplot as plt
from udgsizes.obs.sample import load_sample
from udgsizes.obs.index_colour import Classifier, get_classifier_filename
if __name__ == "__main__":
df = load_sample()
indices = df["n_sersic"].values
colours = df["g_r"].values
c = Classifier()
filename = get_classifier_filename()
c.fit(indices, colours, filename=filename, makeplots=True)
plt.show(block=False)