some old updates

Author: Sceki

README.md

@@ -22,26 +22,19 @@
 
 -----------------------------------------
 [![Build Status](https://github.com/kesslerlib/kessler/workflows/build/badge.svg)](https://github.com/kesslerlib/kessler/actions)
-[![Documentation Status](https://readthedocs.org/projects/kessler/badge/?version=latest)](https://kessler.readthedocs.io/en/latest/?badge=latest)
 [![codecov](https://codecov.io/gh/kesslerlib/kessler/branch/master/graph/badge.svg?token=EQ9CLXD909)](https://codecov.io/gh/kesslerlib/kessler)
 
 Kessler is a Python package for simulation-based inference and machine learning for space collision avoidance and assessment. It is named in honor of NASA scientist [Donald J. Kessler](https://en.wikipedia.org/wiki/Donald_J._Kessler) known for his studies regarding [space debris](https://en.wikipedia.org/wiki/Space_debris) and proposing the [Kessler syndrome](https://en.wikipedia.org/wiki/Kessler_syndrome).
 
-Developed by the [FDL Europe](https://fdleurope.org/) Constellations team in collaboration with [European Space Operations Centre (ESOC)](http://www.esa.int/esoc) of the [European Space Agency (ESA)](http://www.esa.int).
+Initially developed by the [FDL Europe](https://fdleurope.org/) Constellations team in collaboration with [European Space Operations Centre (ESOC)](http://www.esa.int/esoc) of the [European Space Agency (ESA)](http://www.esa.int).
 
 ## Documentation and roadmap
 
 To get started, follow the [documentation](https://kesslerlib.github.io/kessler/) examples.
 
 ## Authors
 
-* Giacomo Acciarini, University of Surrey
-* Francesco Pinto, University of Oxford
-* Francesca Letizia, European Space Agency
-* Chris Bridges, University of Surrey
-* Atılım Güneş Baydin, University of Oxford
-
-Kessler was initially developed by the Constellations team at the Frontier Development Lab (FDL) Europe 2020, a public–private partnership between the European Space Agency (ESA), Trillium Technologies, and University of Oxford.
+Kessler was initiated by the Constellations team at the Frontier Development Lab (FDL) Europe 2020, a public–private partnership between the European Space Agency (ESA), Trillium Technologies, and University of Oxford. The main developer is [Giacomo Acciarini](https://www.esa.int/gsp/ACT/team/giacomo_acciarini/).
 
 Constellations team members: Giacomo Acciarini, Francesco Pinto, Sascha Metz, Sarah Boufelja, Sylvester Kaczmarek, Klaus Merz, José A. Martinez-Heras, Francesca Letizia, Christopher Bridges, Atılım Güneş Baydin
 
@@ -51,7 +44,8 @@ Kessler is distributed under the GNU General Public License version 3. Get in to
 
 ## More info and how to cite
 
-If you would like to learn more about or cite the techniques Kessler uses, please see the following papers:
+If you use `kessler`, we would be grateful if you could star the repository and/or cite our work.
+If you would like to learn more about or cite the techniques `kessler` uses, please see the following papers:
 
 * Giacomo Acciarini, Nicola Baresi, Christopher Bridges, Leonard Felicetti, Stephen Hobbs, Atılım Güneş Baydin. 2023. [“Observation Strategies and Megaconstellations Impact on Current LEO Population.”](https://conference.sdo.esoc.esa.int/proceedings/neosst2/paper/88) In 2nd NEO and Debris Detection Conference.
 ```
@@ -63,7 +57,7 @@ If you would like to learn more about or cite the techniques Kessler uses, pleas
 }
 ```
 * Giacomo Acciarini, Francesco Pinto, Francesca Letizia, José A. Martinez-Heras, Klaus Merz, Christopher Bridges, and Atılım Güneş Baydin. 2021. [“Kessler: a Machine Learning Library for Spacecraft Collision Avoidance.”](https://conference.sdo.esoc.esa.int/proceedings/sdc8/paper/226) In 8th European Conference on Space Debris.
-```
+```bibtex
 @inproceedings{acciarini-2020-kessler,
   title = {Kessler: a Machine Learning Library for Spacecraft Collision Avoidance},
   author = {Acciarini, Giacomo and Pinto, Francesco and Letizia, Francesca and Martinez-Heras, José A. and Merz, Klaus and Bridges, Christopher and Baydin, Atılım Güneş},
@@ -72,7 +66,7 @@ If you would like to learn more about or cite the techniques Kessler uses, pleas
 }
 ```
 * Francesco Pinto, Giacomo Acciarini, Sascha Metz, Sarah Boufelja, Sylvester Kaczmarek, Klaus Merz, José A. Martinez-Heras, Francesca Letizia, Christopher Bridges, and Atılım Güneş Baydin. 2020. “Towards Automated Satellite Conjunction Management with Bayesian Deep Learning.” In AI for Earth Sciences Workshop at NeurIPS 2020, Vancouver, Canada. [arXiv:2012.12450](https://arxiv.org/abs/2012.12450)
-```
+```bibtex
 @inproceedings{pinto-2020-automated,
   title = {Towards Automated Satellite Conjunction Management with Bayesian Deep Learning},
   author = {Pinto, Francesco and Acciarini, Giacomo and Metz, Sascha and Boufelja, Sarah and Kaczmarek, Sylvester and Merz, Klaus and Martinez-Heras, José A. and Letizia, Francesca and Bridges, Christopher and Baydin, Atılım Güneş},
@@ -81,7 +75,7 @@ If you would like to learn more about or cite the techniques Kessler uses, pleas
 }
 ```
 * Giacomo Acciarini, Francesco Pinto, Sascha Metz, Sarah Boufelja, Sylvester Kaczmarek, Klaus Merz, José A. Martinez-Heras, Francesca Letizia, Christopher Bridges, and Atılım Güneş Baydin. 2020. “Spacecraft Collision Risk Assessment with Probabilistic Programming.” In Third Workshop on Machine Learning and the Physical Sciences (NeurIPS 2020), Vancouver, Canada. [arXiv:2012.10260](https://arxiv.org/abs/2012.10260)
-```
+```bibtex
 @inproceedings{acciarini-2020-spacecraft,
   title = {Spacecraft Collision Risk Assessment with Probabilistic Programming},
   author = {Acciarini, Giacomo and Pinto, Francesco and Metz, Sascha and Boufelja, Sarah and Kaczmarek, Sylvester and Merz, Klaus and Martinez-Heras, José A. and Letizia, Francesca and Bridges, Christopher and Baydin, Atılım Güneş},
@@ -92,7 +86,7 @@ If you would like to learn more about or cite the techniques Kessler uses, pleas
 
 ## Installation
 
-To install kessler, do the following:
+To install `kessler` locally, you can do the following:
 
 ```
 git clone https://github.com/kesslerlib/kessler.git

docs/notebooks/trace.pickle

@@ -1,14 +1,13 @@
 # This code is part of Kessler, a machine learning library for spacecraft collision avoidance.
 #
 # Copyright (c) 2020-
-# University of Oxford (Atilim Gunes Baydin <gunes@robots.ox.ac.uk>)
 # Trillium Technologies
-# Giacomo Acciarini
+# University of Oxford
+# Giacomo Acciarini (giacomo.acciarini@gmail.com)
 # and other contributors, see README in root of repository.
 #
 # GNU General Public License version 3. See LICENSE in root of repository.
 
-
 __version__ = '1.0.0'
 
 from .util import seed

kessler/__init__.py

@@ -1,14 +1,13 @@
 # This code is part of Kessler, a machine learning library for spacecraft collision avoidance.
 #
 # Copyright (c) 2020-
-# University of Oxford (Atilim Gunes Baydin <gunes@robots.ox.ac.uk>)
 # Trillium Technologies
-# Giacomo Acciarini
+# University of Oxford
+# Giacomo Acciarini (giacomo.acciarini@gmail.com)
 # and other contributors, see README in root of repository.
 #
 # GNU General Public License version 3. See LICENSE in root of repository.
 
-
 import numpy as np
 import warnings
 import datetime

kessler/cdm.py

@@ -1,9 +1,9 @@
 # This code is part of Kessler, a machine learning library for spacecraft collision avoidance.
 #
 # Copyright (c) 2020-
-# University of Oxford (Atilim Gunes Baydin <gunes@robots.ox.ac.uk>)
 # Trillium Technologies
-# Giacomo Acciarini
+# University of Oxford
+# Giacomo Acciarini (giacomo.acciarini@gmail.com)
 # and other contributors, see README in root of repository.
 #
 # GNU General Public License version 3. See LICENSE in root of repository.

kessler/data.py

@@ -1,14 +1,13 @@
 # This code is part of Kessler, a machine learning library for spacecraft collision avoidance.
 #
 # Copyright (c) 2020-
-# University of Oxford (Atilim Gunes Baydin <gunes@robots.ox.ac.uk>)
 # Trillium Technologies
-# Giacomo Acciarini
+# University of Oxford
+# Giacomo Acciarini (giacomo.acciarini@gmail.com)
 # and other contributors, see README in root of repository.
 #
 # GNU General Public License version 3. See LICENSE in root of repository.
 
-
 import pandas as pd
 import numpy as np
 import matplotlib as mpl

kessler/event.py

@@ -1,3 +1,13 @@
+# This code is part of Kessler, a machine learning library for spacecraft collision avoidance.
+#
+# Copyright (c) 2020-
+# Trillium Technologies
+# University of Oxford
+# Giacomo Acciarini (giacomo.acciarini@gmail.com)
+# and other contributors, see README in root of repository.
+#
+# GNU General Public License version 3. See LICENSE in root of repository.
+
 import dsgp4
 import numpy as np
 import uuid
@@ -8,10 +18,6 @@
 from . import GNSS, Radar, ConjunctionDataMessage, util
 from dsgp4 import TLE
 
-from torch.distributions import constraints
-from torch.distributions.distribution import Distribution
-from torch.distributions.utils import broadcast_all
-
 from pyro.distributions import MixtureSameFamily, Categorical, Uniform, Normal, Bernoulli
 
 def find_conjunction(tr0, 
@@ -45,74 +51,6 @@ def find_conjunction(tr0,
         d_conj = squared_norm[i_conj].sqrt()
         return i_min, d_min, i_conj, d_conj
 
-
-class TruncatedNormal(Distribution):
-    """
-    Truncated Normal distribution with specified lower and upper bounds.
-    This class inherits from the Pyro Distribution class and implements
-    the log probability and sampling methods for a truncated normal distribution.
-
-    Args:
-        loc (``torch.Tensor``): The mean of the normal distribution.
-        scale (``torch.Tensor``): The standard deviation of the normal distribution.
-        low (``torch.Tensor``, optional): The lower bound for truncation. Default is None.
-        high (``torch.Tensor``, optional): The upper bound for truncation. Default is None.
-        validate_args (bool, optional): Whether to validate the arguments. Default is None.
-
-    Attributes:
-        loc (``torch.Tensor``): The mean of the normal distribution.
-        scale (``torch.Tensor``): The standard deviation of the normal distribution.
-        low (``torch.Tensor``): The lower bound for truncation.
-        high (``torch.Tensor``): The upper bound for truncation.
-        base_dist (``Normal``): The base normal distribution.
-
-    Methods:
-        log_prob(value): Computes the log probability of the given value.
-        sample(sample_shape): Samples from the truncated normal distribution.
-    """
-    arg_constraints = {
-        'loc': constraints.real,       # loc can be any real number
-        'scale': constraints.positive, # scale must be positive
-        'low': constraints.real,       # low can be any real number
-        'high': constraints.real       # high can be any real number
-    }
-
-    def __init__(self, loc, scale, low=None, high=None, validate_args=None):
-        # Convert inputs to tensors and handle None values
-        self.loc, self.scale, self.low, self.high = broadcast_all(
-            torch.as_tensor(loc), torch.as_tensor(scale),
-            torch.as_tensor(low) if low is not None else None,
-            torch.as_tensor(high) if high is not None else None
-        )
-        # Validate bounds (low < high)
-        if self.low is not None and self.high is not None:
-            if (self.low >= self.high).any():
-                raise ValueError("Invalid bounds: low must be less than high.")
-        # Create a base Normal distribution
-        self.base_dist = Normal(loc, scale)
-        super().__init__(batch_shape=self.loc.shape, validate_args=validate_args)
-
-    def log_prob(self, value):
-        # Calculate log probability for the normal distribution
-        log_prob = self.base_dist.log_prob(value)
-        # Adjust for truncation at the low bound (if any)
-        if self.low is not None:
-            log_prob -= torch.log(torch.clamp(self.base_dist.cdf(self.low), min=1e-10))
-        # Adjust for truncation at the high bound (if any)
-        if self.high is not None:
-            log_prob -= torch.log(torch.clamp(1 - self.base_dist.cdf(self.high), min=1e-10))
-
-        return log_prob
-
-    def sample(self, sample_shape=torch.Size()):
-        shape = self._extended_shape(sample_shape)
-        rand = torch.rand(shape, dtype=self.loc.dtype, device=self.loc.device)
-        if self.low is not None:
-            rand = rand * (1 - self.base_dist.cdf(self.low)) + self.base_dist.cdf(self.low)
-        if self.high is not None:
-            rand = rand * self.base_dist.cdf(self.high)
-        return self.base_dist.icdf(rand)
-
 def default_prior():
     """
     This function returns a dictionary of TLE elements priors.
@@ -145,7 +83,7 @@ def default_prior():
         mixture_distribution=Categorical(probs=torch.tensor([
             0.12375596165657043, 0.05202080309391022, 0.21220888197422028,
             0.0373813770711422, 0.01674230769276619, 0.5578906536102295])),
-        component_distribution=TruncatedNormal(
+        component_distribution=util.TruncatedNormal(
             low=0.0, high=0.004, loc=torch.tensor([
                 0.0010028142482042313, 0.00017592836171388626, 0.0010926761478185654,
                 0.0003353552892804146, 0.0007777251303195953, 0.001032940074801445]),
@@ -162,7 +100,7 @@ def default_prior():
         mixture_distribution=Categorical(probs=torch.tensor([
             0.5433819890022278, 0.04530993849039078, 0.08378008753061295,
             0.02705608867108822, 0.03350389748811722, 0.2669680118560791])),
-        component_distribution=TruncatedNormal(
+        component_distribution=util.TruncatedNormal(
             low=0.0, high=0.8999999761581421, loc=torch.tensor([
                 0.0028987403493374586, 0.6150050163269043, 0.05085373669862747,
                 0.3420163094997406, 0.7167646288871765, 0.013545362278819084]),
@@ -178,7 +116,7 @@ def default_prior():
             0.028676774352788925, 0.06484941393136978, 0.13786117732524872, 0.0010146398562937975,
             0.047179922461509705, 0.01607278548181057, 0.020023610442876816, 0.06644929945468903,
             0.4442509114742279])),
-        component_distribution=TruncatedNormal(
+        component_distribution=util.TruncatedNormal(
             low=0.0, high=torch.pi, loc=torch.tensor([
                 0.09954200685024261, 1.4393062591552734, 1.736578106880188, 1.0963480472564697,
                 0.48166394233703613, 0.9063634872436523, 1.275956392288208, 2.5208728313446045,

kessler/model.py

@@ -1,9 +1,9 @@
 # This code is part of Kessler, a machine learning library for spacecraft collision avoidance.
 #
 # Copyright (c) 2020-
-# University of Oxford (Atilim Gunes Baydin <gunes@robots.ox.ac.uk>)
 # Trillium Technologies
-# Giacomo Acciarini
+# University of Oxford
+# Giacomo Acciarini (giacomo.acciarini@gmail.com)
 # and other contributors, see README in root of repository.
 #
 # GNU General Public License version 3. See LICENSE in root of repository.

kessler/nn.py

@@ -1,3 +1,13 @@
+# This code is part of Kessler, a machine learning library for spacecraft collision avoidance.
+#
+# Copyright (c) 2020-
+# Trillium Technologies
+# University of Oxford
+# Giacomo Acciarini (giacomo.acciarini@gmail.com)
+# and other contributors, see README in root of repository.
+#
+# GNU General Public License version 3. See LICENSE in root of repository.
+
 import numpy as np
 
 

kessler/observation_model.py

@@ -26,14 +26,15 @@ def read_package_variable(key):
 setup(
     name='kessler',
     version=read_package_variable('__version__'),
-    description='Simulation-based inference and machine learning for space collision assessment and avoidance.',
-    author='ESA FDL Europe Constellations Team',
-    # author_email='',
+    description='Machine learning and simulation-based inference and machine learning for space collision assessment and avoidance.',
+    long_description=open('README.md').read(),
+    author='Giacomo Acciarini',
+    author_email='giacomo.acciarini@gmail.com',
     packages=find_packages(),
     install_requires=['pyro', 'numpy', 'matplotlib', 'torch>=1.5.1', 'dsgp4', 'skyfield>=1.26', 'pandas'],
     extras_require={'dev': ['pytest', 'coverage', 'pytest-xdist']},
-    # url='https://github.com/kessler/kessler',
+    url='https://kesslerlib.github.io/kessler/',
     classifiers=['License :: OSI Approved :: BSD License', 'Programming Language :: Python :: 3'],
     license='BSD'
-    # keywords='',
+    keywords='Spacecraft Collision Avoidance Kessler Machine Learning Artificial Intelligence Probabilistic Programming',
 )