Merge pull request janelia-flyem#82 from jni/speedup

Speed up RAG building and agglomeration:
- 30x speedup in RAG building
- 5x speedup in flat learning
- 6x speedup in agglomerative learning
- 7x speedup in test segmentation
- And 30% reduction in RAM usage by the RAG.

Key changes:
- New fast RAG building method based on dilation and erosion of labels.
- New row-expandable CSR format for the contingency matrix.
- New sparselol-based boundary extents.
- New fast testing framework.
- New benchmarking file.
- Don't update edges when they are unchanged by a merge.

.travis.yml

@@ -6,7 +6,7 @@ virtualenv:
   system_site_packages: false
 install:
   # all installing is now handled by conda as it is faster and more robust
-  - wget http://repo.continuum.io/miniconda/Miniconda-3.4.2-Linux-x86_64.sh -O miniconda.sh;
+  - wget http://repo.continuum.io/miniconda/Miniconda-latest-Linux-x86_64.sh -O miniconda.sh;
   - bash miniconda.sh -b -p $HOME/miniconda
   - export PATH="$HOME/miniconda/bin:$PATH"
   - hash -r

asv.conf.json

@@ -0,0 +1,134 @@
+{
+    // The version of the config file format.  Do not change, unless
+    // you know what you are doing.
+    "version": 1,
+
+    // The name of the project being benchmarked
+    "project": "gala",
+
+    // The project's homepage
+    "project_url": "http://gala.readthedocs.io/",
+
+    // The URL or local path of the source code repository for the
+    // project being benchmarked
+    "repo": ".",
+
+    // List of branches to benchmark. If not provided, defaults to "master"
+    // (for git) or "tip" (for mercurial).
+    "branches": ["master", "speedup"], // for git
+    // "branches": ["tip"],    // for mercurial
+
+    // The DVCS being used.  If not set, it will be automatically
+    // determined from "repo" by looking at the protocol in the URL
+    // (if remote), or by looking for special directories, such as
+    // ".git" (if local).
+    // "dvcs": "git",
+
+    // The tool to use to create environments.  May be "conda",
+    // "virtualenv" or other value depending on the plugins in use.
+    // If missing or the empty string, the tool will be automatically
+    // determined by looking for tools on the PATH environment
+    // variable.
+    "environment_type": "conda",
+
+    // the base URL to show a commit for the project.
+    "show_commit_url": "http://github.com/jni/gala/commit/",
+
+    // The Pythons you'd like to test against.  If not provided, defaults
+    // to the current version of Python used to run `asv`.
+    // "pythons": ["2.7", "3.3"],
+
+    // The matrix of dependencies to test.  Each key is the name of a
+    // package (in PyPI) and the values are version numbers.  An empty
+    // list or empty string indicates to just test against the default
+    // (latest) version. null indicates that the package is to not be
+    // installed.
+    //
+    "matrix": {
+        "environment_type": "conda",
+        "python": "3.5",
+        "numpy": [],
+        "scipy": [],
+        "pip+numpydoc": [],
+        "networkx": [],
+        "h5py": [],
+        "cython": [],
+        "pip+viridis": [],
+        "pyzmq": [],
+        "scikit-learn": [],
+        "scikit-image": [],
+        "pytest": [],
+        "setuptools": [],
+        "coverage": []
+    },
+
+    // Combinations of libraries/python versions can be excluded/included
+    // from the set to test. Each entry is a dictionary containing additional
+    // key-value pairs to include/exclude.
+    //
+    // An exclude entry excludes entries where all values match. The
+    // values are regexps that should match the whole string.
+    //
+    // An include entry adds an environment. Only the packages listed
+    // are installed. The 'python' key is required. The exclude rules
+    // do not apply to includes.
+    //
+    // In addition to package names, the following keys are available:
+    //
+    // - python
+    //     Python version, as in the *pythons* variable above.
+    // - environment_type
+    //     Environment type, as above.
+    // - sys_platform
+    //     Platform, as in sys.platform. Possible values for the common
+    //     cases: 'linux2', 'win32', 'cygwin', 'darwin'.
+    //
+    // "exclude": [
+    //     {"python": "3.2", "sys_platform": "win32"}, // skip py3.2 on windows
+    //     {"environment_type": "conda", "six": null}, // don't run without six on conda
+    // ],
+    //
+    // "include": [
+    //     // additional env for python2.7
+    //     {"python": "2.7", "numpy": "1.8"},
+    //     // additional env if run on windows+conda
+    //     {"platform": "win32", "environment_type": "conda", "python": "2.7", "libpython": ""},
+    // ],
+
+    // The directory (relative to the current directory) that benchmarks are
+    // stored in.  If not provided, defaults to "benchmarks"
+    // "benchmark_dir": "benchmarks",
+
+    // The directory (relative to the current directory) to cache the Python
+    // environments in.  If not provided, defaults to "env"
+    "env_dir": ".asv/env",
+
+    // The directory (relative to the current directory) that raw benchmark
+    // results are stored in.  If not provided, defaults to "results".
+    "results_dir": ".asv/results",
+
+    // The directory (relative to the current directory) that the html tree
+    // should be written to.  If not provided, defaults to "html".
+    "html_dir": ".asv/html"
+
+    // The number of characters to retain in the commit hashes.
+    // "hash_length": 8,
+
+    // `asv` will cache wheels of the recent builds in each
+    // environment, making them faster to install next time.  This is
+    // number of builds to keep, per environment.
+    // "wheel_cache_size": 0
+
+    // The commits after which the regression search in `asv publish`
+    // should start looking for regressions. Dictionary whose keys are
+    // regexps matching to benchmark names, and values corresponding to
+    // the commit (exclusive) after which to start looking for
+    // regressions.  The default is to start from the first commit
+    // with results. If the commit is `null`, regression detection is
+    // skipped for the matching benchmark.
+    //
+    // "regressions_first_commits": {
+    //    "some_benchmark": "352cdf",  // Consider regressions only after this commit
+    //    "another_benchmark": null,   // Skip regression detection altogether
+    // }
+}

benchmarks/bench_gala.py

@@ -0,0 +1,136 @@
+import os
+
+from contextlib import contextmanager
+from collections import OrderedDict
+
+import numpy as np
+
+from gala import imio, features, agglo, classify
+from asv.extern.asizeof import asizeof
+
+
+rundir = os.path.dirname(__file__)
+## dd: the data directory
+dd = os.path.abspath(os.path.join(rundir, '../tests/example-data'))
+
+
+from time import process_time
+
+
+@contextmanager
+def timer():
+    time = []
+    t0 = process_time()
+    yield time
+    t1 = process_time()
+    time.append(t1 - t0)
+
+
+em = features.default.paper_em()
+
+
+def trdata():
+    wstr = imio.read_h5_stack(os.path.join(dd, 'train-ws.lzf.h5'))
+    prtr = imio.read_h5_stack(os.path.join(dd, 'train-p1.lzf.h5'))
+    gttr = imio.read_h5_stack(os.path.join(dd, 'train-gt.lzf.h5'))
+    return wstr, prtr, gttr
+
+
+def tsdata():
+    wsts = imio.read_h5_stack(os.path.join(dd, 'test-ws.lzf.h5'))
+    prts = imio.read_h5_stack(os.path.join(dd, 'test-p1.lzf.h5'))
+    gtts = imio.read_h5_stack(os.path.join(dd, 'test-gt.lzf.h5'))
+    return wsts, prts, gtts
+
+
+def trgraph():
+    ws, pr, ts = trdata()
+    g = agglo.Rag(ws, pr)
+    return g
+
+
+def tsgraph():
+    ws, pr, ts = tsdata()
+    g = agglo.Rag(ws, pr, feature_manager=em)
+    return g
+
+
+def trexamples():
+    gt = imio.read_h5_stack(os.path.join(dd, 'train-gt.lzf.h5'))
+    g = trgraph()
+    (X, y, w, e), _ = g.learn_agglomerate(gt, em, min_num_epochs=5)
+    y = y[:, 0]
+    return X, y
+
+
+def classifier():
+    X, y = trexamples()
+    rf = classify.DefaultRandomForest()
+    rf.fit(X, y)
+    return rf
+
+
+def policy():
+    rf = classify.DefaultRandomForest()
+    cl = agglo.classifier_probability(em, rf)
+    return cl
+
+
+def tsgraph_queue():
+    g = tsgraph()
+    cl = policy()
+    g.merge_priority_function = cl
+    g.rebuild_merge_queue()
+    return g
+
+def bench_suite():
+    times = OrderedDict()
+    memory = OrderedDict()
+    wstr, prtr, gttr = trdata()
+    with timer() as t_build_rag:
+        g = agglo.Rag(wstr, prtr)
+    times['build RAG'] = t_build_rag[0]
+    memory['base RAG'] = asizeof(g)
+    with timer() as t_features:
+        g.set_feature_manager(em)
+    times['build feature caches'] = t_features[0]
+    memory['feature caches'] = asizeof(g) - memory['base RAG']
+    with timer() as t_flat:
+        _ignore = g.learn_flat(gttr, em)
+    times['learn flat'] = t_flat[0]
+    with timer() as t_gala:
+        (X, y, w, e), allepochs = g.learn_agglomerate(gttr, em,
+                                                      min_num_epochs=5)
+        y = y[:, 0]  # ignore rand-sign and vi-sign schemes
+    memory['training data'] = asizeof((X, y, w, e))
+    times['learn agglo'] = t_gala[0]
+    with timer() as t_train_classifier:
+        cl = classify.DefaultRandomForest()
+        cl.fit(X, y)
+    times['classifier training'] = t_train_classifier[0]
+    memory['classifier training'] = asizeof(cl)
+    policy = agglo.classifier_probability(em, cl)
+    wsts, prts, gtts = tsdata()
+    gtest = agglo.Rag(wsts, prts, merge_priority_function=policy,
+                      feature_manager=em)
+    with timer() as t_segment:
+        gtest.agglomerate(np.inf)
+    times['segment test volume'] = t_segment[0]
+    memory['segment test volume'] = asizeof(gtest)
+    return times, memory
+
+
+def print_bench_results(times=None, memory=None):
+    if times is not None:
+        print('Timing results:')
+        for key in times:
+            print('--- ', key, times[key])
+    if memory is not None:
+        print('Memory results:')
+        for key in memory:
+            print('--- ', key, '%.3f MB' % (memory[key] / 1e6))
+
+
+if __name__ == '__main__':
+    times, memory = bench_suite()
+    print_bench_results(times, memory)