AMR-utils is a python package for working with AMR data, with tools for reading AMRs and alignments, performing graph operations, and displaying and visualizing AMR data. I wrote this package to store operations that I often need or find useful when doing research with AMRs. This code is written and maintained by Austin Blodgett.
- Load AMRs from a file or directory, with support for multiple formats
- Load AMR alignments, with support for LDC, JAMR, and ISI alignment formats
- A simple class for accessing AMR nodes, edges, tokens, etc.
- Graph operations for operating on AMR data
- Tools for AMR Visualization
- Convert AMR graphs to Latex (using the tikz library)
- Display AMR strings as HTML, with overridable display settings for nodes, tokens, and edges
- AMR Diff, Display differnces between AMRs as HTML
- Display AMR Alignments as HTML
- Python 3.6 or higher
- PENMAN library
git clone https://github.com//ablodge/amr-utils
pip install penman
pip install ./amr-utils
If you have a question that isn't answered by this document, please check the Wiki.
- A small excerpt of code is taken from smatch for AMR-to-AMR alignment in the AMR Diff tool, so that results from AMR Diff are directly related to the smatch score.
The class AMR_Reader can be used to load AMRs or AMR alignments from a number of different formats including LDC, JAMR, and ISI. An AMR_Reader can be used as follows.
from amr_utils.amr_readers import AMR_Reader
reader = AMR_Reader()
amrs = reader.load(amr_file, remove_wiki=True)
AMRs must be separated by empty lines, but otherwise can take various formats. Simplified:
# Dogs chase cats.
(c/chase-01 :ARG0 (d/dog)
:ARG1 (c2/cat))
JAMR-style graph metdata format:
# ::id 1
# ::tok Dogs chase cats.
# ::node c chase-01
# ::node d dog
# ::node c2 cat
# ::root c chase-01
# ::edge chase-01 ARG0 dog c d
# ::edge chase-01 ARG1 cat c c2
(c/chase-01 :ARG0 (d/dog)
:ARG1 (c2/cat))
AMR Alignments can also be loaded from different formats:
- LDC:
# ::alignments 0-1.1 1-1 1-1.1.r 1-1.2.r 2-1.2 - JAMR:
# ::alignments 0-1|0.0 1-2|0 2-3|0.1 - ISI:
(c/chase-01~e.1 :ARG0~e.1 (d/dog~e.0) :ARG1~e.1 (c2/cat~e.2))
Just set the parameter output_alignments to True.
from amr_utils.amr_readers import AMR_Reader
reader = AMR_Reader()
amrs, alignments = reader.load(amr_file, remove_wiki=True, output_alignments=True)
By default, AMR_Reader uses the LDC/ISI style of node ids where 1.n is the nth child of the root with indices starting at 1.
Any alignments are automatically converted to this format for data consistency.
The package includes tools for converting AMR alignments from and to JSON like the following.
[{'type':'isi', 'tokens':[0], 'nodes':['1.1'], 'edges':[]},
{'type':'isi', 'tokens':[1], 'nodes':['1'], 'edges':[['1',':ARG0','1.1'],['1',':ARG1','1.2']]},
{'type':'isi', 'tokens':[2], 'nodes':['1.2'], 'edges':[]},
]
The advantages of using JSON are:
- Easy to load and save (No need to write a special script for reading some esoteric format)
- Can store additional information in a
typeto distinguish different types of alignments - Can easily store multiple sets of alignments separately, without needing to modify an AMR file. That makes it easy to compare different sets of alignments.
To read alignments from a JSON file do:
reader = AMR_Reader()
alignments = reader.load_alignments_from_json(alignments_file)
To save alignments to a JSON file do:
reader = AMR_Reader()
reader.save_alignments_to_json(alignments_file, alignments)
AMR-utils includes tools for visualizing AMRs and AMR aligments. See the wiki for more detail.
AMR-utils allows you to read AMRs from a text file and output them as latex diagrams, such as the following.
The default coloring assigns blue to each node, but the parameter assign_color can be used to assign colors using a function. To change a color by hand, just rewrite \node[red] as \node[purple], etc.
Run as follows:
python style.py --latex -f [input file] [output file]
Add these lines to your latex file:
\usepackage{tikz}
\usetikzlibrary{shapes}
AMR-utils allows you to read AMRs from a text file and output them as html. You can look in style.css for an example of styling.
Run as follows:
python style.py --html -f [input file] [output file]
AMR Diff is a tool for comparing two files of AMRs. The tool uses AMR-to-AMR alignment from smatch to find the differences between pairs of AMRs which contribute to a lower smatch score. AMR Diff is useful for detailed error analysis of AMR parsers. The display includes highlighted differences and mouse-over description text explanation of the error.
Run as follows:
python amr_diff.py [amr file1] [amr file2] [output file]
AMR-utils also includes a tool for displaying alignments in an easy-to-read format, with highlights and mouse-over description text of which tokens/nodes/edges are aligned.
Run as follows:
python display_alignments.py [amr file] [alignment file] [output file]
You can import graph operations from graph_utils.py:
from amr_utils.graph_utils import get_subgraph, is_rooted_dag, breadth_first_nodes, \
breadth_first_edges, depth_first_nodes, depth_first_edges, \
get_shortest_path, get_connected_components
Functions in graph_utils.py allow you to
- Traverse AMR nodes or edges in depth-first or breadth-first order
- Retrieve an AMR subgraph
- Test an AMR or sub-AMR for DAG structure
- Get the shortest path between two nodes
- Seperate a subset of AMR nodes into connected components