Source code

Download source archive ↓

resources/META-INF/galaxy-operator.json

1{"categories": ["Galaxy", "Statistics", "Data analysis"], "description": "Fit binary logistic regression with ridge-stabilized IRLS.", "documentation": {"example_job": "name: Fit binary logistic regression with ridge-stabilized IRLS.\nsteps:\n- id: read\n  operator: read_item@1.0\n  inputs: {}\n  params:\n    uri: context:/galaxy/extended/logistic_regression.txt\n    view: bytes\n- id: process\n  operator: galaxy_logistic_regression@1.0\n  inputs:\n    source: read\n  params: {}\noutputs:\n  result: process\n", "inputs": {"source": "UTF-8 bytes stream. Binary response0/1 with both classes, <=16 standardized predictors, <=100 damped Newton steps; rejects singularity and nonconvergence. Coefficients only, not native VIF diagnostics. Strict headerless numeric TSV; finite values of magnitude <=1e12, <=10000 rows,128 columns,8 MiB characters. Missing/non-numeric data rejected; no R subprocesses or PDF diagnostics."}, "outputs": {"result": "Fit binary logistic regression with ridge-stabilized IRLS. Binary response0/1 with both classes, <=16 standardized predictors, <=100 damped Newton steps; rejects singularity and nonconvergence. Coefficients only, not native VIF diagnostics. Strict headerless numeric TSV; finite values of magnitude <=1e12, <=10000 rows,128 columns,8 MiB characters. Missing/non-numeric data rejected; no R subprocesses or PDF diagnostics."}, "parameters": {"predictor_columns": {"description": "Unique numeric predictors, response excluded.", "example": [1], "help": "Unique numeric predictors, response excluded. Binary response0/1 with both classes, <=16 standardized predictors, <=100 damped Newton steps; rejects singularity and nonconvergence. Coefficients only, not native VIF diagnostics. Strict headerless numeric TSV; finite values of magnitude <=1e12, <=10000 rows,128 columns,8 MiB characters. Missing/non-numeric data rejected; no R subprocesses or PDF diagnostics."}, "response_column": {"description": "One-based numeric response column.", "example": 2, "help": "One-based numeric response column. Binary response0/1 with both classes, <=16 standardized predictors, <=100 damped Newton steps; rejects singularity and nonconvergence. Coefficients only, not native VIF diagnostics. Strict headerless numeric TSV; finite values of magnitude <=1e12, <=10000 rows,128 columns,8 MiB characters. Missing/non-numeric data rejected; no R subprocesses or PDF diagnostics."}, "ridge_x1000": {"description": "L2 penalty multiplied by1000, applied to standardized predictors, not intercept.", "example": 1, "help": "L2 penalty multiplied by1000, applied to standardized predictors, not intercept. Binary response0/1 with both classes, <=16 standardized predictors, <=100 damped Newton steps; rejects singularity and nonconvergence. Coefficients only, not native VIF diagnostics. Strict headerless numeric TSV; finite values of magnitude <=1e12, <=10000 rows,128 columns,8 MiB characters. Missing/non-numeric data rejected; no R subprocesses or PDF diagnostics."}}}, "entrypoint": "xstream.galaxy.LogisticRegression", "fixture": {"data": "-1\t0\n-1\t1\n1\t0\n1\t1\n", "expected": "term\tcoefficient\nintercept\t0.000000\ncolumn_1\t0.000000\n", "params": {}}, "icon": {"color": "#5265ad", "path": "M4 3h12l4 4v14H4zM8 11h8M8 15h8m-8 4h5"}, "implementation": "StatisticalScience", "inputs": {"source": {"formats": [], "representations": ["bytes"]}}, "name": "galaxy_logistic_regression", "parameters": {"additionalProperties": false, "properties": {"predictor_columns": {"default": [1], "description": "Unique numeric predictors, response excluded.", "items": {"maximum": 128, "minimum": 1, "type": "integer"}, "maxItems": 16, "type": "array"}, "response_column": {"default": 2, "description": "One-based numeric response column.", "maximum": 128, "minimum": 1, "type": "integer"}, "ridge_x1000": {"default": 1, "description": "L2 penalty multiplied by1000, applied to standardized predictors, not intercept.", "maximum": 1000000, "minimum": 1, "type": "integer"}}, "required": [], "type": "object"}, "params": {}, "scope": "Binary response0/1 with both classes, <=16 standardized predictors, <=100 damped Newton steps; rejects singularity and nonconvergence. Coefficients only, not native VIF diagnostics. Strict headerless numeric TSV; finite values of magnitude <=1e12, <=10000 rows,128 columns,8 MiB characters. Missing/non-numeric data rejected; no R subprocesses or PDF diagnostics.", "upstream": {"path": "tools/rpy_statistics_collection/logistic_regression_vif.xml", "repository": "https://github.com/bgruening/galaxytools"}, "version": "1.0"}