merge from master (c3771da2) · Commits · aflab / astrogeology / Autocnet

.travis.yml

+3 −1

Original line number	Diff line number	Diff line
		@@ -9,6 +9,7 @@ python:
		- "3.5"

		before_install:

		install:
		# We do this conditionally because it saves us some downloading if the
		# version is the same.
		@@ -26,11 +27,12 @@ install:
		- conda info -a

		# Create a virtual env and install dependencies
		- conda create -y -q -n test-env python=$TRAVIS_PYTHON_VERSION nose gdal numpy scipy pandas
		- conda create -y -q -n test-env python=$TRAVIS_PYTHON_VERSION nose gdal numpy pillow scipy pandas networkx
		# Activate the env
		- source activate test-env

		# Install the non-conda packages if required, requirements.txt duplicates are ignored
		- conda install -c https://conda.binstar.org/menpo opencv3
		- conda install -c osgeo proj4
		- pip install -r requirements.txt
		- pip install coverage

autocnet/control/control.py

+2 −0

Original line number	Diff line number	Diff line
		@@ -2,6 +2,8 @@ from time import gmtime, strftime

		import pandas as pd

		POINT_TYPE = 2
		MEASURE_TYPE = 2

		class CSeries(pd.Series):
		"""

autocnet/fileio/io_gdal.py

+1 −0

Original line number	Diff line number	Diff line
		@@ -350,6 +350,7 @@ class GeoDataset(object):
		xextent, yextent).astype(dtype)
		return array


		def array_to_raster(array, file_name, projection=None,
		geotransform=None, outformat='GTiff',
		ndv=None):

autocnet/fileio/tests/io_controlnetwork.py

deleted100644 → 0

+0 −0

Empty file deleted.

autocnet/graph/network.py

+97 −7

Original line number	Diff line number	Diff line
		from collections import Hashable
		from networkx import DiGraph

		import networkx as nx
		import pandas as pd

		from autocnet.control.control import C, POINT_TYPE, MEASURE_TYPE
		from autocnet.fileio import io_json
		class CandidateGraph(DiGraph):



		class CandidateGraph(nx.Graph):
		"""
		A NetworkX derived directed graph to store candidate overlap images.

		@@ -12,8 +18,6 @@ class CandidateGraph(DiGraph):
		Attributes
		----------
		"""
		#TODO: This would be better with composition, and then dispatch the
		# network X calls to the graph object.

		def __init__(self,args, *kwargs):
		super(CandidateGraph, self).__init__(args, *kwargs)
		@@ -36,6 +40,7 @@ class CandidateGraph(DiGraph):

		Parameters
		==========

		outputfile : str
		PATH where the JSON will be written
		"""
		@@ -44,18 +49,103 @@ class CandidateGraph(DiGraph):
		adjacency_dict[n] = self.neighbors(n)
		io_json.write_json(adjacency_dict, outputfile)

		def add_matches(self, source_node, matches):
		"""

		Adds match data to a node and attributes the data to the
		appropriate edges, e.g. if A-B have a match, edge A-B is attributes
		with the pandas dataframe.

		Parameters
		----------
		source_node : str
		The identifier for the node

		matches : dataframe
		The pandas dataframe containing the matches
		"""

		#TODO: This really belongs in an outlier detection matcher class, not here.
		# Remove erroneous self neighbors
		matches = matches.loc[matches['matched_to'] != source_node]

		groups = matches.groupby('matched_to')
		for destination_node, group in groups:
		try:
		edge = self[source_node][destination_node]
		except:
		edge = self[destination_node][source_node]

		if 'matches' in edge.keys():
		df = edge['matches']
		edge['matches'] = pd.merge(df, matches, left_on='queryIdx', right_on='trainIdx')
		else:
		edge['matches'] = matches

		def to_cnet(self):
		"""

		Returns
		-------
		cnet : C
		A control network object
		"""
		data = []
		point_ids = []
		serials = []
		for source, destination, attributes in self.edges_iter(data=True):

		for i, row in attributes['matches'].iterrows():

		# Get the source and destination x,y coordinates for matches
		source_idx = row['queryIdx_x']
		source_keypoints = (self.node[source]['keypoints'][source_idx].pt[0],
		self.node[source]['keypoints'][source_idx].pt[1])

		destination_idx = row['queryIdx_y']
		destination_keypoints = (self.node[destination]['keypoints'][destination_idx].pt[0],
		self.node[destination]['keypoints'][destination_idx].pt[1])

		data.append(source_keypoints)
		data.append(destination_keypoints)
		serials.append(source)
		serials.append(destination)
		point_ids.append(i)
		point_ids.append(i)

		point_types = [2] * len(point_ids)
		measure_types = [2] * len(point_ids)
		multi_index = pd.MultiIndex.from_tuples(list(zip(point_ids,point_types,
		serials, measure_types)))

		columns = ['x', 'y']
		cnet = C(data, index=multi_index, columns=columns)

		# TODO: This method assumes a 2 image match and should be generalized to build an n-image C object

		return cnet

		@classmethod
		def from_adjacency(cls, inputfile):
		"""
		Instantiate the class using an adjacency list

		Parameters
		==========
		----------
		inputfile : str
		The input file containing the graph representation

		Returns
		-------
		: object
		A Network graph object

		Examples
		--------
		>>> from autocnet.examples import get_path
		>>> inputfile = get_path('adjacency.json')
		>>> candidate_graph = network.CandidateGraph.from_adjacency(inputfile)
		"""
		#TODO: This is better as a generic reader that tries drivers until
		# a valid dict is returned.
		adjacency_dict = io_json.read_json(inputfile)
		return cls(adjacency_dict)