{"id":"W2947203165","doi":"10.1109/escience.2019.00023","title":"Evaluation of Pilot Jobs for Apache Spark Applications on HPC Clusters","year":2019,"lang":"en","type":"preprint","venue":"","topic":"Cloud Computing and Resource Management","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Concordia University","funders":"Western Canada Research Grid; Compute Canada","keywords":"SPARK (programming language); Computer science; Big data; Scheduling (production processes); Supercomputer; Operating system; Debugging; Software deployment; Database; Engineering","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.007927509,0.001762882,0.001007513,0.0009753337,0.001169023,0.001386842,0.002402218,0.0008528198,0.002287019],"category_scores_gemma":[0.02298508,0.000726633,0.0005917244,0.001446464,0.001077048,0.001814607,0.001130891,0.001496609,0.0008082665],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001216548,"about_ca_system_score_gemma":0.002006786,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0117034,"about_ca_topic_score_gemma":0.006006524,"domain_scores_codex":[0.9956297,0.001426872,0.000394698,0.0005738639,0.001278306,0.0006965704],"domain_scores_gemma":[0.9726589,0.01439497,0.0009226894,0.002958398,0.006067039,0.00299799],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.04815396,0.01159857,0.06450772,0.003975102,0.001106479,0.002373137,0.003409937,0.3866366,0.1482552,0.008426119,0.08036896,0.2411883],"study_design_scores_gemma":[0.002616001,0.01018653,0.06352965,0.000140791,0.0001684864,0.0003878372,0.001694613,0.8221349,0.07727912,0.002418783,0.01919775,0.0002455429],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9680915,0.000572006,0.01110267,0.0003543572,0.0005408042,0.0005846451,0.001172435,0.01128212,0.006299455],"genre_scores_gemma":[0.9686571,0.0002127034,0.02528093,0.0001165241,0.00006455132,0.0002267814,0.002853093,0.001092858,0.001495495],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.0117034,"threshold_uncertainty_score":0.04192519,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09134819555802973,"score_gpt":0.3208827718341735,"score_spread":0.2295345762761437,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}