\n",
"\n",
"**This is a fixed-text formatted version of a Jupyter notebook**\n",
"\n",
"- Try online [![Binder](https://static.mybinder.org/badge.svg)](https://mybinder.org/v2/gh/gammapy/gammapy-webpage/v0.18.2?urlpath=lab/tree/ring_background.ipynb)\n",
"- You can contribute with your own notebooks in this\n",
"[GitHub repository](https://github.com/gammapy/gammapy/tree/master/docs/tutorials).\n",
"- **Source files:**\n",
"[ring_background.ipynb](../_static/notebooks/ring_background.ipynb) |\n",
"[ring_background.py](../_static/notebooks/ring_background.py)\n",
"
\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Ring Background Estimation\n",
"\n",
"## Context:\n",
"One of the challenges of IACT analysis is accounting for the large residual hadronic emission. An excess map, assumed to be a map of only gamma-ray events, requires a good estimate of the background. However, in the absence of a solid template bkg model it is not possible to obtain reliable background model a priori. It was often found necessary in classical cherenkov astronomy to perform a local renormalization of the existing templates, usually with a ring kernel. This assumes that most of the events are background and requires to have an exclusion mask to remove regions with bright signal from the estimation. To read more about this method, see [here.](https://arxiv.org/abs/astro-ph/0610959)\n",
"\n",
"## Objective:\n",
"Create an excess (gamma-ray events) map of MSH 15-52 as well as a significance map to determine how solid the signal is.\n",
"\n",
"## Proposed approach:\n",
"\n",
"The analysis workflow is roughly\n",
" - Compute the sky maps keeping each observation separately using the `Analysis` class\n",
" - Estimate the background using the `RingBackgroundMaker`\n",
" - Compute the correlated excess and significance maps using the `CorrelatedExcessMapEstimator`\n",
" \n",
"The normalised background thus obtained can be used for general modelling and fitting."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Setup\n",
"As usual, we'll start with some general imports..."
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"%matplotlib inline\n",
"import astropy.units as u\n",
"from astropy.coordinates import SkyCoord\n",
"import matplotlib.pyplot as plt\n",
"import numpy as np\n",
"from regions import CircleSkyRegion\n",
"from scipy.stats import norm\n",
"\n",
"import logging\n",
"\n",
"log = logging.getLogger(__name__)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"Now let's import gammapy specific classes and functions"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"from gammapy.analysis import Analysis, AnalysisConfig\n",
"from gammapy.makers import RingBackgroundMaker\n",
"from gammapy.estimators import ExcessMapEstimator\n",
"from gammapy.maps import Map\n",
"from gammapy.datasets import MapDatasetOnOff"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Creating the config file\n",
"Now, we create a config file for out analysis. You may load this from disc if you have a pre-defined config file.\n",
"\n",
"In this example, we will use a few HESS runs on the pulsar wind nebula, MSH 1552"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
"# source_pos = SkyCoord.from_name(\"MSH 15-52\")\n",
"source_pos = SkyCoord(228.32, -59.08, unit=\"deg\")"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
"config = AnalysisConfig()\n",
"# Select observations - 2.5 degrees from the source position\n",
"config.observations.datastore = \"$GAMMAPY_DATA/hess-dl3-dr1/\"\n",
"config.observations.obs_cone = {\n",
" \"frame\": \"icrs\",\n",
" \"lon\": source_pos.ra,\n",
" \"lat\": source_pos.dec,\n",
" \"radius\": 2.5 * u.deg,\n",
"}"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [],
"source": [
"config.datasets.type = \"3d\"\n",
"config.datasets.geom.wcs.skydir = {\n",
" \"lon\": source_pos.ra,\n",
" \"lat\": source_pos.dec,\n",
" \"frame\": \"icrs\",\n",
"} # The WCS geometry - centered on MSH 15-52\n",
"config.datasets.geom.wcs.fov = {\"width\": \"3 deg\", \"height\": \"3 deg\"}\n",
"config.datasets.geom.wcs.binsize = \"0.02 deg\"\n",
"\n",
"# The FoV radius to use for cutouts\n",
"config.datasets.geom.selection.offset_max = 3.5 * u.deg\n",
"\n",
"# We now fix the energy axis for the counts map - (the reconstructed energy binning)\n",
"config.datasets.geom.axes.energy.min = \"0.5 TeV\"\n",
"config.datasets.geom.axes.energy.max = \"5 TeV\"\n",
"config.datasets.geom.axes.energy.nbins = 10\n",
"\n",
"# We need to extract the ring for each observation separately, hence, no stacking at this stage\n",
"config.datasets.stack = False"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"AnalysisConfig\n",
"\n",
" general:\n",
" log: {level: info, filename: null, filemode: null, format: null, datefmt: null}\n",
" outdir: .\n",
" observations:\n",
" datastore: $GAMMAPY_DATA/hess-dl3-dr1\n",
" obs_ids: []\n",
" obs_file: null\n",
" obs_cone: {frame: icrs, lon: 228.32 deg, lat: -59.08 deg, radius: 2.5 deg}\n",
" obs_time: {start: null, stop: null}\n",
" datasets:\n",
" type: 3d\n",
" stack: false\n",
" geom:\n",
" wcs:\n",
" skydir: {frame: icrs, lon: 228.32 deg, lat: -59.08 deg}\n",
" binsize: 0.02 deg\n",
" fov: {width: 3.0 deg, height: 3.0 deg}\n",
" binsize_irf: 0.2 deg\n",
" selection: {offset_max: 3.5 deg}\n",
" axes:\n",
" energy: {min: 0.5 TeV, max: 5.0 TeV, nbins: 10}\n",
" energy_true: {min: 0.1 TeV, max: 10.0 TeV, nbins: 30}\n",
" map_selection: [counts, exposure, background, psf, edisp]\n",
" background:\n",
" method: null\n",
" exclusion: null\n",
" parameters: {}\n",
" safe_mask:\n",
" methods: [aeff-default]\n",
" parameters: {}\n",
" on_region: {frame: null, lon: null, lat: null, radius: null}\n",
" containment_correction: true\n",
" fit:\n",
" fit_range: {min: 0.1 TeV, max: 10.0 TeV}\n",
" flux_points:\n",
" energy: {min: 0.1 TeV, max: 10.0 TeV, nbins: 30}\n",
" source: source\n",
" parameters: {}\n",
" \n"
]
}
],
"source": [
"print(config)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Getting the reduced dataset\n",
"We now use the config file to do the initial data reduction which will then be used for a ring extraction"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"Setting logging config: {'level': 'INFO', 'filename': None, 'filemode': None, 'format': None, 'datefmt': None}\n",
"Fetching observations.\n",
"Number of selected observations: 20\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"CPU times: user 205 ms, sys: 2.98 ms, total: 208 ms\n",
"Wall time: 208 ms\n"
]
}
],
"source": [
"%%time\n",
"# create the config\n",
"analysis = Analysis(config)\n",
"\n",
"# for this specific case,w e do not need fine bins in true energy\n",
"analysis.config.datasets.geom.axes.energy_true = (\n",
" analysis.config.datasets.geom.axes.energy\n",
")\n",
"\n",
"# `First get the required observations\n",
"analysis.get_observations()"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"AnalysisConfig\n",
"\n",
" general:\n",
" log: {level: INFO, filename: null, filemode: null, format: null, datefmt: null}\n",
" outdir: .\n",
" observations:\n",
" datastore: $GAMMAPY_DATA/hess-dl3-dr1\n",
" obs_ids: []\n",
" obs_file: null\n",
" obs_cone: {frame: icrs, lon: 228.32 deg, lat: -59.08 deg, radius: 2.5 deg}\n",
" obs_time: {start: null, stop: null}\n",
" datasets:\n",
" type: 3d\n",
" stack: false\n",
" geom:\n",
" wcs:\n",
" skydir: {frame: icrs, lon: 228.32 deg, lat: -59.08 deg}\n",
" binsize: 0.02 deg\n",
" fov: {width: 3.0 deg, height: 3.0 deg}\n",
" binsize_irf: 0.2 deg\n",
" selection: {offset_max: 3.5 deg}\n",
" axes:\n",
" energy: {min: 0.5 TeV, max: 5.0 TeV, nbins: 10}\n",
" energy_true: {min: 0.5 TeV, max: 5.0 TeV, nbins: 10}\n",
" map_selection: [counts, exposure, background, psf, edisp]\n",
" background:\n",
" method: null\n",
" exclusion: null\n",
" parameters: {}\n",
" safe_mask:\n",
" methods: [aeff-default]\n",
" parameters: {}\n",
" on_region: {frame: null, lon: null, lat: null, radius: null}\n",
" containment_correction: true\n",
" fit:\n",
" fit_range: {min: 0.1 TeV, max: 10.0 TeV}\n",
" flux_points:\n",
" energy: {min: 0.1 TeV, max: 10.0 TeV, nbins: 30}\n",
" source: source\n",
" parameters: {}\n",
" \n"
]
}
],
"source": [
"print(analysis.config)"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"Creating geometry.\n",
"Creating datasets.\n",
"No background maker set for 3d analysis. Check configuration.\n",
"Processing observation 20365\n",
"Processing observation 20366\n",
"Processing observation 20367\n",
"Processing observation 20368\n",
"Processing observation 20136\n",
"Processing observation 20137\n",
"Processing observation 20151\n",
"Processing observation 20282\n",
"Processing observation 20283\n",
"Processing observation 20301\n",
"Processing observation 20302\n",
"Processing observation 20303\n",
"Processing observation 20322\n",
"Processing observation 20323\n",
"Processing observation 20324\n",
"Processing observation 20325\n",
"Processing observation 20343\n",
"Processing observation 20344\n",
"Processing observation 20345\n",
"Processing observation 20346\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"CPU times: user 13.4 s, sys: 1.19 s, total: 14.6 s\n",
"Wall time: 14.7 s\n"
]
}
],
"source": [
"%%time\n",
"# Data extraction\n",
"analysis.get_datasets()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Extracting the ring background\n",
"\n",
"Since the ring background is extracted from real off events, we need to use the wstat statistics in this case. For this, we will use the `MapDatasetOnOFF` and the `RingBackgroundMaker` classes."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Create exclusion mask\n",
"First, we need to create an exclusion mask on the known sources. In this case, we need to mask only `MSH 15-52` but this depends on the sources present in our field of view."
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "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\n",
"text/plain": [
"