del tree_normalizer print "finished training tree-norm SVM:", svm_tree.get_objective() wdk = shogun_factory.create_kernel(data.examples, param) wdk.set_normalizer(normalizer) wdk.init_normalizer() print "--->",wdk.get_normalizer().get_name() svm = SVMLight(cost, wdk, lab) svm.set_linadd_enabled(False) svm.set_batch_computation_enabled(False) svm.train() print "finished training manually set SVM:", svm.get_objective() alphas_tree = svm_tree.get_alphas() alphas = svm.get_alphas() assert(len(alphas_tree)==len(alphas)) for i in xrange(len(alphas)): assert(abs(alphas_tree[i] - alphas[i]) < 0.0001) print "success: all alphas are the same"
def test_data(): ################################################################## # select MSS ################################################################## mss = expenv.MultiSplitSet.get(379) ################################################################## # data ################################################################## # fetch data instance_set = mss.get_train_data(-1) # prepare data data = PreparedMultitaskData(instance_set, shuffle=True) # set parameters param = Options() param.kernel = "WeightedDegreeStringKernel" param.wdk_degree = 4 param.cost = 1.0 param.transform = 1.0 param.id = 666 param.freeze() ################################################################## # taxonomy ################################################################## taxonomy = shogun_factory.create_taxonomy(mss.taxonomy.data) support = numpy.linspace(0, 100, 4) distances = [[0, 1, 2, 2], [1, 0, 2, 2], [2, 2, 0, 1], [2, 2, 1, 0]] # create tree normalizer tree_normalizer = MultitaskKernelPlifNormalizer(support, data.task_vector_names) task_names = data.get_task_names() FACTOR = 1.0 # init gamma matrix gammas = numpy.zeros((data.get_num_tasks(), data.get_num_tasks())) for t1_name in task_names: for t2_name in task_names: similarity = taxonomy.compute_node_similarity(taxonomy.get_id(t1_name), taxonomy.get_id(t2_name)) gammas[data.name_to_id(t1_name), data.name_to_id(t2_name)] = similarity helper.save("/tmp/gammas", gammas) gammas = gammas * FACTOR cost = param.cost * numpy.sqrt(FACTOR) print gammas ########## # regular normalizer normalizer = MultitaskKernelNormalizer(data.task_vector_nums) for t1_name in task_names: for t2_name in task_names: similarity = gammas[data.name_to_id(t1_name), data.name_to_id(t2_name)] normalizer.set_task_similarity(data.name_to_id(t1_name), data.name_to_id(t2_name), similarity) ################################################################## # Train SVMs ################################################################## # create shogun objects wdk_tree = shogun_factory.create_kernel(data.examples, param) lab = shogun_factory.create_labels(data.labels) wdk_tree.set_normalizer(tree_normalizer) wdk_tree.init_normalizer() print "--->",wdk_tree.get_normalizer().get_name() svm_tree = SVMLight(cost, wdk_tree, lab) svm_tree.set_linadd_enabled(False) svm_tree.set_batch_computation_enabled(False) svm_tree.train() del wdk_tree del tree_normalizer print "finished training tree-norm SVM:", svm_tree.get_objective() wdk = shogun_factory.create_kernel(data.examples, param) wdk.set_normalizer(normalizer) wdk.init_normalizer() print "--->",wdk.get_normalizer().get_name() svm = SVMLight(cost, wdk, lab) svm.set_linadd_enabled(False) svm.set_batch_computation_enabled(False) svm.train() print "finished training manually set SVM:", svm.get_objective() alphas_tree = svm_tree.get_alphas() alphas = svm.get_alphas() assert(len(alphas_tree)==len(alphas)) for i in xrange(len(alphas)): assert(abs(alphas_tree[i] - alphas[i]) < 0.0001) print "success: all alphas are the same"
def solver_mtk_shogun(C, all_xt, all_lt, task_indicator, M, L, eps, target_obj): """ implementation using multitask kernel """ xt = numpy.array(all_xt) lt = numpy.array(all_lt) tt = numpy.array(task_indicator, dtype=numpy.int32) tsm = numpy.array(M) print "task_sim:", tsm num_tasks = L.shape[0] # sanity checks assert len(xt) == len(lt) == len(tt) assert M.shape == L.shape assert num_tasks == len(set(tt)) # set up shogun objects if type(xt[0]) == numpy.string_: feat = StringCharFeatures(DNA) xt = [str(a) for a in xt] feat.set_features(xt) base_kernel = WeightedDegreeStringKernel(feat, feat, 8) else: feat = RealFeatures(xt.T) base_kernel = LinearKernel(feat, feat) lab = Labels(lt) # set up normalizer normalizer = MultitaskKernelNormalizer(tt.tolist()) for i in xrange(num_tasks): for j in xrange(num_tasks): normalizer.set_task_similarity(i, j, M[i, j]) print "num of unique tasks: ", normalizer.get_num_unique_tasks( task_indicator) # set up kernel base_kernel.set_cache_size(2000) base_kernel.set_normalizer(normalizer) base_kernel.init_normalizer() # set up svm svm = SVMLight() #LibSVM() svm.set_epsilon(eps) #print "reducing num threads to one" #svm.parallel.set_num_threads(1) #print "using one thread" # how often do we like to compute objective etc svm.set_record_interval(0) svm.set_target_objective(target_obj) svm.set_linadd_enabled(False) svm.set_batch_computation_enabled(False) svm.io.set_loglevel(MSG_DEBUG) #SET THREADS TO 1 svm.set_C(C, C) svm.set_bias_enabled(False) # prepare for training svm.set_labels(lab) svm.set_kernel(base_kernel) # train svm svm.train() train_times = svm.get_training_times() objectives = [-obj for obj in svm.get_dual_objectives()] if False: # get model parameters sv_idx = svm.get_support_vectors() sparse_alphas = svm.get_alphas() assert len(sv_idx) == len(sparse_alphas) # compute dense alpha (remove label) alphas = numpy.zeros(len(xt)) for id_sparse, id_dense in enumerate(sv_idx): alphas[id_dense] = sparse_alphas[id_sparse] * lt[id_dense] # print alphas W = alphas_to_w(alphas, xt, lt, task_indicator, M) primal_obj = compute_primal_objective( W.reshape(W.shape[0] * W.shape[1]), C, all_xt, all_lt, task_indicator, L) objectives.append(primal_obj) train_times.append(train_times[-1] + 100) return objectives, train_times
for idx in xrange(10): #f = (-numpy.ones(N)-2)*numpy.random.randn() print "############################" print "############################" print "" print "f:", f print "\n" svm.set_linear_term(numpy.double(f)) svm.train() sv_idx = svm.get_support_vectors() alphas = svm.get_alphas() alphas_full = numpy.zeros(N) alphas_full[sv_idx] = alphas alphas_full = alphas_full * y print "svmlight objective:", svm.get_objective() print "svmlight alphas:", numpy.array(alphas_full[0:5]) external_objective = 0.0 for j in xrange(N): external_objective += alphas_full[j] * f[j]
def test_data(): ################################################################## # select MSS ################################################################## mss = expenv.MultiSplitSet.get(379) ################################################################## # data ################################################################## # fetch data instance_set = mss.get_train_data(-1) # prepare data data = PreparedMultitaskData(instance_set, shuffle=True) # set parameters param = Options() param.kernel = "WeightedDegreeStringKernel" param.wdk_degree = 4 param.cost = 1.0 param.transform = 1.0 param.id = 666 param.freeze() ################################################################## # taxonomy ################################################################## taxonomy = shogun_factory.create_taxonomy(mss.taxonomy.data) support = numpy.linspace(0, 100, 4) distances = [[0, 1, 2, 2], [1, 0, 2, 2], [2, 2, 0, 1], [2, 2, 1, 0]] # create tree normalizer tree_normalizer = MultitaskKernelPlifNormalizer(support, data.task_vector_names) task_names = data.get_task_names() FACTOR = 1.0 # init gamma matrix gammas = numpy.zeros((data.get_num_tasks(), data.get_num_tasks())) for t1_name in task_names: for t2_name in task_names: similarity = taxonomy.compute_node_similarity( taxonomy.get_id(t1_name), taxonomy.get_id(t2_name)) gammas[data.name_to_id(t1_name), data.name_to_id(t2_name)] = similarity helper.save("/tmp/gammas", gammas) gammas = gammas * FACTOR cost = param.cost * numpy.sqrt(FACTOR) print gammas ########## # regular normalizer normalizer = MultitaskKernelNormalizer(data.task_vector_nums) for t1_name in task_names: for t2_name in task_names: similarity = gammas[data.name_to_id(t1_name), data.name_to_id(t2_name)] normalizer.set_task_similarity(data.name_to_id(t1_name), data.name_to_id(t2_name), similarity) ################################################################## # Train SVMs ################################################################## # create shogun objects wdk_tree = shogun_factory.create_kernel(data.examples, param) lab = shogun_factory.create_labels(data.labels) wdk_tree.set_normalizer(tree_normalizer) wdk_tree.init_normalizer() print "--->", wdk_tree.get_normalizer().get_name() svm_tree = SVMLight(cost, wdk_tree, lab) svm_tree.set_linadd_enabled(False) svm_tree.set_batch_computation_enabled(False) svm_tree.train() del wdk_tree del tree_normalizer print "finished training tree-norm SVM:", svm_tree.get_objective() wdk = shogun_factory.create_kernel(data.examples, param) wdk.set_normalizer(normalizer) wdk.init_normalizer() print "--->", wdk.get_normalizer().get_name() svm = SVMLight(cost, wdk, lab) svm.set_linadd_enabled(False) svm.set_batch_computation_enabled(False) svm.train() print "finished training manually set SVM:", svm.get_objective() alphas_tree = svm_tree.get_alphas() alphas = svm.get_alphas() assert (len(alphas_tree) == len(alphas)) for i in xrange(len(alphas)): assert (abs(alphas_tree[i] - alphas[i]) < 0.0001) print "success: all alphas are the same"
svm_tree.train() del wdk_tree del tree_normalizer print "finished training tree-norm SVM:", svm_tree.get_objective() wdk = shogun_factory.create_kernel(data.examples, param) wdk.set_normalizer(normalizer) wdk.init_normalizer() print "--->", wdk.get_normalizer().get_name() svm = SVMLight(cost, wdk, lab) svm.set_linadd_enabled(False) svm.set_batch_computation_enabled(False) svm.train() print "finished training manually set SVM:", svm.get_objective() alphas_tree = svm_tree.get_alphas() alphas = svm.get_alphas() assert (len(alphas_tree) == len(alphas)) for i in xrange(len(alphas)): assert (abs(alphas_tree[i] - alphas[i]) < 0.0001) print "success: all alphas are the same"