Version 0.7.5

What's Changed

• Add survival analysis support by @selcukorkmaz in #58
• correcting regression naming convention by @selcukorkmaz in #59
• adding survival task by @selcukorkmaz in #60
• fix survival label check in fastml by @selcukorkmaz in #61
• Handle MICE imputation for survival tasks by @selcukorkmaz in #62
• Use aorsf engine for survival random forest by @selcukorkmaz in #63
• Ensure survival random forest loads required packages by @selcukorkmaz in #64
• Handle non-data frame survival predictions by @selcukorkmaz in #65
• Fix survival risk predictions for native models by @selcukorkmaz in #66
• Fix coxnet survival predictions by @selcukorkmaz in #67
• Fix coxnet risk predictions by @selcukorkmaz in #68
• Remove coxnet survival support by @selcukorkmaz in #69
• Compute integrated Brier scores for survival models by @selcukorkmaz in #70
• Add survival support for Brier scoring and royston model by @selcukorkmaz in #71
• Fix survreg Brier score computation and royston-parmar training by @selcukorkmaz in #72
• Fix survival workflows for stratified and Royston-Parmar models by @selcukorkmaz in #73
• Fix log-rank calculation for time-varying Cox by @selcukorkmaz in #74
• Revamp survival metric reporting by @selcukorkmaz in #75
• Fix survival summary metric alignment by @selcukorkmaz in #76
• Allow customizing survival summary metrics output by @selcukorkmaz in #77
• Fix summary CI formatting when intervals unavailable by @selcukorkmaz in #78
• Improve survival summary defaults and status handling by @selcukorkmaz in #79
• Remove confusion matrix warning for non-classification tasks by @selcukorkmaz in #80
• Improve survival summary Brier time handling by @selcukorkmaz in #81
• Improve survival model summaries in summary.fastml by @selcukorkmaz in #82
• Handle stratified Cox summary details by @selcukorkmaz in #83
• Fix stratified Cox handling of dummy strata columns by @selcukorkmaz in #84
• Improve stratified Cox handling and summary by @selcukorkmaz in #85
• Respect metric direction when selecting best model by @selcukorkmaz in #86
• Ensure h2o ROC AUC metric retains metric name by @selcukorkmaz in #87
• Prevent outcome leakage during advanced imputation by @selcukorkmaz in #88
• Handle time_varying_cox in summary details by @selcukorkmaz in #89
• Add Royston-Parmar model details to summary by @selcukorkmaz in #90
• Support rstpm2 survival metrics and summary by @selcukorkmaz in #91
• Improve Royston-Parmar summary metrics by @selcukorkmaz in #92
• Add engine_params argument for engine-level options by @selcukorkmaz in #93
• Fix rstpm2 summary call extraction and labels by @selcukorkmaz in #94
• Improve rstpm2 summaries by @selcukorkmaz in #95
• Sanitize Royston-Parmar summary call output by @selcukorkmaz in #96
• Add penalized Cox regression support by @selcukorkmaz in #97
• Add xgboost survival engines and prediction helpers by @selcukorkmaz in #98
• Update xgboost survival handling for Cox and AFT by @selcukorkmaz in #99
• Fix xgboost AFT survival post-processing by @selcukorkmaz in #100
• Remove xgboost support from survival task by @selcukorkmaz in #101
• Add flexsurv parametric survival engines by @selcukorkmaz in #102
• Improve survival metric handling and flexsurv distribution mapping by @selcukorkmaz in #103
• Handle flexsurv summary row outputs for survival metrics by @selcukorkmaz in #104
• Ensure flexsurv survival metrics and piecewise exponential setup by @selcukorkmaz in #105
• Add flexsurv handler for parametric survival metrics by @selcukorkmaz in #106
• Fix flexsurv survival predictions and metrics by @selcukorkmaz in #107
• Handle flexsurv survival summaries for parametric models by @selcukorkmaz in #108
• Fix flexsurv survival extraction for IBS metric by @selcukorkmaz in #109
• Improve flexsurv survival predictions by @selcukorkmaz in #110
• Add explicit S3 annotations for fastml methods by @selcukorkmaz in #111
• Ensure native survival predictor helper is globally available by @selcukorkmaz in #112
• Fix NA survival predictions for flexsurv parametric models by @selcukorkmaz in #113
• Handle flexsurvreg survival predictions by @selcukorkmaz in #114
• Improve flexsurv survival probability extraction by @selcukorkmaz in #115
• Vectorize flexsurv survival matrix fallback by @selcukorkmaz in #116
• Fix flexsurvreg prediction data handling by @selcukorkmaz in #117
• Vectorize flexsurv survival matrix computation by @selcukorkmaz in #118
• Fix flexsurv summary lookup for survival predictions by @selcukorkmaz in #119
• Fix flexsurv summary newdata selection by @selcukorkmaz in #120
• Use optimized flexsurv survival matrix helper by @selcukorkmaz in #121
• Ensure survival interpolation is right-continuous by @selcukorkmaz in #122
• Handle flexsurvreg survival time calculation by @selcukorkmaz in #123
• Fix flexsurvreg survival time computation by @selcukorkmaz in #124
• Fix flexsurvreg quantile computation in process_model by @selcukorkmaz in #125
• Handle flexsurv quantile calls for parametric survival by @selcukorkmaz in #126
• Fix flexsurvreg survival time computation by @selcukorkmaz in #127
• Calculate survival medians from survival curves by @selcukorkmaz in #128
• Prevent early exit when computing survival times by @selcukorkmaz in #129
• Implement custom piecewise exponential survival distribution by @selcukorkmaz in #130
• Stabilize piecewise exponential survival fitting by @selcukorkmaz in #131
• Fix piecewise exponential rate extraction for flexsurv integration by @selcukorkmaz in #132
• Handle singular Hessian when fitting flexsurv models by @selcukorkmaz in #133
• Add xgboost AFT support for survival task by @selcukorkmaz in #134
• Ensure XGBoost survival risk scores fall back to AFT mean by @selcukorkmaz in #135
• Fix XGBoost AFT survival probability calculation by @selcukorkmaz in #136
• Fix xgboost AFT manual block to use margin predictions by @selcukorkmaz in #137
• Fix XGBoost AFT survival prediction by @selcukorkmaz in #138
• Align XGBoost survival predictors with training features by @selcukorkmaz in #139
• Fix summary details extraction for xgboost survival models by @selcukorkmaz in #140
• Handle survival labels in predict.fastml by @selcukorkmaz in #141
• Fix random forest tuning grid naming collision by @selcukorkmaz in #142
• Fix native survival risk predictions to use preprocessed data by @selcukorkmaz in #143
• Guard against data leakage during resampling by @selcukorkmaz in #144
• Enforce guarded preprocessing across resampling workflows by @selcukorkmaz in #146
• Allow survival resampling for parsnip survival engines by @selcukorkmaz in #147
• Enable fold-wise advanced imputation during resampling by @selcukorkmaz in #148
• Harden recipe validation and sandbox custom preprocessing by @selcukorkmaz in #149
• Improve recipe leakage checks and sandbox defaults by @selcukorkmaz in #150
• Ensure sandbox fetch retrieves exported functions by @selcukorkmaz in #151
• Extend train_models resampling options by @selcukorkmaz in #152
• Fix block count calculation in blocked CV by @selcukorkmaz in #153
• Implement grouped cross-validation helper by @selcukorkmaz in #154
• Handle novel categorical levels during prediction by @selcukorkmaz in #155
• Handle missing categorical levels during prediction by @selcukorkmaz in #156
• Prevent QDA rank deficiency on imputed data by @selcukorkmaz in #157
• Use sparsediscrim engine for quadratic discriminant by @selcukorkmaz in #158
• Add helper-based blocked and rolling resampling by @selcukorkmaz in #159
• Implement unbiased nested cross-validation by @selcukorkmaz in #160
• Integrate resampling metadata into training workflow by @selcukorkmaz in #161
• Add targeted resampling strategy unit tests by @selcukorkmaz in #162
• Enhance JSS manuscript prompt with title guidance by @selcukorkmaz in #163
• Add JSS manuscript for fastml by @selcukorkmaz in #164
• Devel by @selcukorkmaz in #165
• Devel by @selcukorkmaz in #166

Full Changelog: release_0.6.2...release_0.7.5