{:group-id "scicloj", :artifact-id "metamorph.ml", :version "0.7.10", :analysis {"clj" ({:name scicloj.metamorph.ml, :publics ({:name default-loss-fn, :file "scicloj/metamorph/ml.clj", :line 707, :arglists ([dataset]), :doc "Given a datset which must have exactly 1 inference target column return a default\nloss fn. If column is categorical, loss is tech.v3.ml.loss/classification-loss, else\nthe loss is tech.v3.ml.loss/mae (mean average error).", :type :var} {:name default-result-dissoc-in-fn, :file "scicloj/metamorph/ml.clj", :line 282, :arglists ([result]), :type :var} {:name default-result-dissoc-in-seq, :file "scicloj/metamorph/ml.clj", :line 263, :type :var} {:name define-model!, :file "scicloj/metamorph/ml.clj", :line 509, :arglists ([model-kwd train-fn predict-fn {:keys [hyperparameters thaw-fn explain-fn loglik-fn options documentation unsupervised?], :as opts}]), :doc "Create a model definition.  An ml model is a function that takes a dataset and an\noptions map and returns a model.  A model is something that, combined with a dataset,\nproduces a inferred dataset.", :type :var} {:name ensemble-pipe, :file "scicloj/metamorph/ml/ensemble.clj", :line 18, :arglists ([pipes]), :doc "Creates an ensemble pipeline function out of various pipelines. The different predictions\nget combined via majority voting.\nCan be used in the same way as any other pipeline.", :type :var} {:name evaluate-pipelines, :file "scicloj/metamorph/ml.clj", :line 314, :arglists ([pipe-fn-or-decl-seq train-test-split-seq metric-fn loss-or-accuracy options] [pipe-fn-seq train-test-split-seq metric-fn loss-or-accuracy]), :doc "Evaluates the performance of a seq of metamorph pipelines, which are suposed to have a model as last step under key :model,\nwhich behaves correctly  in mode :fit and  :transform. The function `scicloj.metamorph.ml/model` is such function behaving correctly.\n\n This function calculates the accuracy or loss, given as `metric-fn` of each pipeline in `pipeline-fn-seq` using all the train-test splits\ngiven in  `train-test-split-seq`.\n\n It runs the pipelines  in mode  :fit and in mode :transform for each pipeline-fn in `pipe-fn-seq` for each split in `train-test-split-seq`.\n\n The function returns a seq of seqs of evaluation results per pipe-fn per train-test split.\n Each of the evaluation results is a context map, which is specified in the malli schema attached to this function.\n\n * `pipe-fn-or-decl-seq` need to be  sequence of pipeline functions or pipline declarations which follow the metamorph approach.\n    These type of functions get produced typically by calling `scicloj.metamorph/pipeline`. Documentation is here:\n\n * `train-test-split-seq` need to be a sequence of maps containing the  train and test dataset (being tech.ml.dataset) at keys :train and :test.\n  `tablecloth.api/split->seq` produces such splits. Supervised models require both keys (:train and :test), while unsupervised models only use :train\n\n * `metric-fn` Metric function to use. Typically comming from `tech.v3.ml.loss`. For supervised models the metric-fn receives the trueth\n    and predicted values and should return a single double number.  For unsupervised models the function receives the fitted ctx\n    and should return a singel double number as well. This metric will be used to sort and eventualy filter the result, depending on the options\n    (:return-best-pipeline-only   and :return-best-crossvalidation-only). The notion of `best` comes from metric-fn combined with loss-and-accuracy\n\n\n * `loss-or-accuracy` If the metric-fn is a loss or accuracy calculation. Can be :loss or :accuracy. Decided the notion of `best` model.\n    In case of :loss pipelines with lower metric are better, in case of :accuracy pipelines with higher value are better.\n\n* `options` map controls some mainly performance related parameters. These function can potentialy result in a large ammount of data,\n  able to bring the JVM into out-of-memory. We can control how many details the function returns by the following parameter: \n   The default are quite aggresive in removing details, and this can be tweaked further into more or less details via:\n   \n\n\n     * `:return-best-pipeline-only` - Only return information of the best performing pipeline. Default is true.\n     * `:return-best-crossvalidation-only` - Only return information of the best crossvalidation (per pipeline returned). Default is true.\n     * `:map-fn` - Controls parallelism, so if we use map (:map) , pmap (:pmap) or :mapv to map over different pipelines. Default :pmap\n     * `:evaluation-handler-fn` - Gets called once with the complete result of an individual pipeline evaluation.\n         It can be used to adapt the data returned for each evaluation and / or to make side effects using\n         the evaluatio data.\n         The result of this function is taken as evaluation result. It need to  contain as a minumum this 2 key paths:\n         [:train-transform :metric]\n         [:test-transform :metric]\n         All other evalution data can be removed, if desired.\n\n         It can be used for side effects as well, like experiment tracking on disk.\n         The passed in evaluation result is a map with all information on the current evaluation, including the datasets used.\n\n         The default handler function is:  `scicloj.metamorph.ml/default-result-dissoc--in-fn` which removes the often large\n         model object and the training data.\n         `identity` can be use to get all evaluation data.\n         `scicloj.metamorph.ml/result-dissoc-in-seq--all` reduces even more agressively.\n\n\n\n     * `:other-metrices` Specifies other metrices to be calculated during evaluation\n\n This function expects as well the ground truth of the target variable into\n a specific key in the context at key `:model :scicloj.metamorph.ml/target-ds`\n See here for the simplest way to set this up: https://github.com/behrica/metamorph.ml/blob/main/README.md\n The function [[scicloj.ml.metamorph/model]] does this correctly.\n", :type :var} {:name explain, :file "scicloj/metamorph/ml.clj", :line 694, :arglists ([model & [options]]), :doc "Explain (if possible) an ml model.  A model explanation is a model-specific map\nof data that usually indicates some level of mapping between features and importance", :type :var} {:name get-categorical-maps, :file "scicloj/metamorph/ml.clj", :line 28, :arglists ([ds]), :type :var} {:name hyperparameters, :file "scicloj/metamorph/ml.clj", :line 565, :arglists ([model-kwd]), :doc "Get the hyperparameters for this model definition\n", :type :var} {:name loglik, :file "scicloj/metamorph/ml.clj", :line 684, :arglists ([model y yhat]), :type :var} {:name model, :file "scicloj/metamorph/ml.clj", :line 724, :arglists ([options]), :doc "Executes a machine learning model in train/predict (depending on :mode)\nfrom the `metamorph.ml` model registry.\n\nThe model is passed between both invocation via the shared context ctx in a\nkey (a step indentifier) which is passed in key `:metamorph/id` and guarantied to be unique for each\npipeline step.\n\nThe function writes and reads into this common context key.\n\nOptions:\n- `:model-type` - Keyword for the model to use\n\nFurther options get passed to `train` functions and are model specific.\n\nSee here for an overview for the models build into scicloj.ml:\n\n\nhttps://scicloj.github.io/scicloj.ml-tutorials/userguide-models.html\n\nOther libraries might contribute other models,\nwhich are documented as part of the library.\n\n\nmetamorph                            | .\n-------------------------------------|----------------------------------------------------------------------------\nBehaviour in mode :fit               | Calls `scicloj.metamorph.ml/train` using data in `:metamorph/data` and `options`and stores trained model in ctx under key in `:metamorph/id`\nBehaviour in mode :transform         | Reads trained model from ctx and calls `scicloj.metamorph.ml/predict` with the model in $id and data in `:metamorph/data`\nReads keys from ctx                  | In mode `:transform` : Reads trained model to use for prediction from key in `:metamorph/id`.\nWrites keys to ctx                   | In mode `:fit` : Stores trained model in key $id and writes feature-ds and target-ds before prediction into ctx at `:scicloj.metamorph.ml/feature-ds` /`:scicloj.metamorph.ml/target-ds`\n\n\n\n\nSee as well:\n\n* `scicloj.metamorph.ml/train`\n* `scicloj.metamorph.ml/predict`\n\n", :type :var} {:name model-definition-names, :file "scicloj/metamorph/ml.clj", :line 550, :arglists ([]), :doc "Return a list of all registered model defintion names.\n", :type :var} {:name model-definitions*, :file "scicloj/metamorph/ml.clj", :line 506, :doc "Map of model kwd to model definition\n", :type :var} {:name options->model-def, :file "scicloj/metamorph/ml.clj", :line 556, :arglists ([options]), :doc "Return the model definition that corresponse to the :model-type option\n", :type :var} {:name predict, :file "scicloj/metamorph/ml.clj", :line 658, :arglists ([dataset model]), :doc "Predict returns a dataset with only the predictions in it.\n\n* For regression, a single column dataset is returned with the column named after the\n  target\n* For classification, a dataset is returned with a float64 column for each target\n  value and values that describe the probability distribution.", :type :var} {:name result-dissoc-in-seq--all, :file "scicloj/metamorph/ml.clj", :line 295, :type :var} {:name result-dissoc-in-seq--all-fn, :file "scicloj/metamorph/ml.clj", :line 310, :arglists ([result]), :type :var} {:name result-dissoc-in-seq--ctxs, :file "scicloj/metamorph/ml.clj", :line 286, :type :var} {:name result-dissoc-in-seq-ctx-fn, :file "scicloj/metamorph/ml.clj", :line 291, :arglists ([result]), :type :var} {:name score, :file "scicloj/metamorph/ml.clj", :line 58, :arglists ([predictions-ds trueth-ds target-column-name metric-fn other-metrices]), :type :var} {:name thaw-model, :file "scicloj/metamorph/ml.clj", :line 610, :arglists ([model {:keys [thaw-fn], :as opts}] [model]), :doc "Thaw a model.  Model's returned from train may be 'frozen' meaning a 'thaw'\noperation is needed in order to use the model.  This happens for you during predict\nbut you may also cached the 'thawed' model on the model map under the\n':thawed-model'  keyword in order to do fast predictions on small datasets.", :type :var} {:name train, :file "scicloj/metamorph/ml.clj", :line 571, :arglists ([dataset options]), :doc "Given a dataset and an options map produce a model.  The model-type keyword in the\noptions map selects which model definition to use to train the model.  Returns a map\ncontaining at least:\n\n\n* `:model-data` - the result of that definitions's train-fn.\n* `:options` - the options passed in.\n* `:id` - new randomly generated UUID.\n* `:feature-columns` - vector of column names.\n* `:target-columns` - vector of column names.", :type :var})} {:name scicloj.metamorph.ml.categorical, :publics ({:name transform-one-hot, :file "scicloj/metamorph/ml/categorical.clj", :line 68, :arglists ([column-selector strategy] [column-selector strategy options]), :doc "Transformer which mapps categorical variables to numbers. Each value of the\ncolumn gets its won column in one-hot-encoding.\n\nTo handle different levls of a variable between train an test data, three\nstrategies are available:\n\n* `:full`  The levels are retrieved from a dataset at key :metamorph.ml/full-ds in the context\n* `:independent`  One-hot columns are fitted and transformed independently for train and test  data\n* `:fit` The mapping fitted in mode :fit is used in :transform, and it is assumed that all levels are present in the data during :fit\n\n`options` can be:\n*  `:table-args` allows to specify the precise mapping as a sequence of pairs of [val idx] or as a sorted seq of values.\n*  `:result-datatype`  Datatype of the one-hot-mapping column\n\n", :type :var})} {:name scicloj.metamorph.ml.classification, :publics ({:name confusion-map, :file "scicloj/metamorph/ml/classification.clj", :line 15, :arglists ([predicted-labels labels normalize] [predicted-labels labels]), :doc "Creates a confusion-matrix in map form. Can be either as raw counts or normalized.\n`normalized` when :all (default) it is normalized\n                  :none otherwise\n ", :type :var} {:name confusion-map->ds, :file "scicloj/metamorph/ml/classification.clj", :line 45, :arglists ([conf-matrix-map]), :doc "Converts teh confusion-matrix map obtained via `confusion-mao` into a dataset representation\n", :type :var})} {:name scicloj.metamorph.ml.ensemble, :publics ({:name ensemble-pipe, :file "scicloj/metamorph/ml/ensemble.clj", :line 18, :arglists ([pipes]), :doc "Creates an ensemble pipeline function out of various pipelines. The different predictions\nget combined via majority voting.\nCan be used in the same way as any other pipeline.", :type :var})} {:name scicloj.metamorph.ml.evaluation-handler, :publics ({:name example-nippy-handler, :file "scicloj/metamorph/ml/evaluation_handler.clj", :line 97, :arglists ([files output-dir result-reduce-fn]), :type :var} {:name file->topforms-with-metadata, :file "scicloj/metamorph/ml/evaluation_handler.clj", :line 11, :arglists ([path]), :type :var} {:name fn-symbol->code, :file "scicloj/metamorph/ml/evaluation_handler.clj", :line 54, :arglists ([fn-symbol pipe-ns pipeline-source-file]), :type :var} {:name fns->code-list, :file "scicloj/metamorph/ml/evaluation_handler.clj", :line 34, :arglists ([pipeline-source-file]), :type :var} {:name fns->code-map, :file "scicloj/metamorph/ml/evaluation_handler.clj", :line 48, :arglists ([pipeline-source-file]), :type :var} {:name get-classpath, :file "scicloj/metamorph/ml/evaluation_handler.clj", :line 77, :arglists ([]), :type :var} {:name get-code, :file "scicloj/metamorph/ml/evaluation_handler.clj", :line 62, :arglists ([symbol pipe-ns pipeline-source-file]), :type :var} {:name get-fn-sources, :file "scicloj/metamorph/ml/evaluation_handler.clj", :line 82, :arglists ([qualified-pipe-decl pipe-ns pipeline-source-file]), :type :var} {:name get-source-information, :file "scicloj/metamorph/ml/evaluation_handler.clj", :line 93, :arglists ([qualified-pipe-decl pipe-ns pipeline-source-file]), :type :var} {:name qualify-keywords, :file "scicloj/metamorph/ml/evaluation_handler.clj", :line 112, :arglists ([pipe-decl pipe-ns]), :type :var} {:name qualify-pipelines, :file "scicloj/metamorph/ml/evaluation_handler.clj", :line 132, :arglists ([pipe-decls pipe-ns]), :type :var} {:name resolve-keyword, :file "scicloj/metamorph/ml/evaluation_handler.clj", :line 19, :arglists ([k pipe-ns]), :doc "Interpret keyword or list as a symbol and try to resolve it.\n", :type :var})} {:name scicloj.metamorph.ml.gridsearch, :publics ({:name categorical, :file "scicloj/metamorph/ml/gridsearch.clj", :line 65, :arglists ([value-vec]), :doc "Given a vector a categorical values create a gridsearch definition.\n", :type :var} {:name linear, :file "scicloj/metamorph/ml/gridsearch.clj", :line 45, :arglists ([start end n-steps res-dtype-or-space] [start end n-steps] [start end]), :doc "Create a gridsearch definition which does a linear search.\n\n* res-dtype-or-space map be either a datatype keyword or a vector\n  of categorical values.", :type :var} {:name sobol-gridsearch, :file "scicloj/metamorph/ml/gridsearch.clj", :line 109, :arglists ([opt-map start-idx] [opt-map]), :doc "Given an map of key->values where some of the values are gridsearch definitions\n  produce a sequence of fully defined maps.\n\n\n```clojure\nuser> (require '[tech.v3.ml.gridsearch :as ml-gs])\nnil\nuser> (def opt-map  {:a (ml-gs/categorical [:a :b :c])\n                     :b (ml-gs/linear 0.01 1 10)\n                     :c :not-searched})\nuser> opt-map\n{:a\n {:tech.v3.ml.gridsearch/type :linear,\n  :start 0.0,\n  :end 2.0,\n  :n-steps 3,\n  :result-space [:a :b :c]}\n  ...\n\nuser> (ml-gs/sobol-gridsearch opt-map)\n({:a :b, :b 0.56, :c :not-searched}\n {:a :c, :b 0.22999999999999998, :c :not-searched}\n {:a :b, :b 0.78, :c :not-searched}\n...\n```\n  ", :type :var}), :doc "Gridsearching as defined by create a map with gridsearch definitions\nfor its values and then gridsearching which produces a sequence of full\ndefined maps.\n\n\nThe initial default implementation uses the sobol sequence."} {:name scicloj.metamorph.ml.learning-curve, :publics ({:name learning-curve, :file "scicloj/metamorph/ml/learning_curve.clj", :line 8, :arglists ([ds pipe-fn train-sizes {:keys [k metric-fn loss-or-accuracy]}]), :doc "Generates a learnining curve.\n\nThe functions splits  the dataset  in a fixed size test set\nand increasingly larger  training sets. A model is trained at each\nstep and evaluated.\n\nReturns dataset with the learninig curve data.\n\n`dataset` the TMD dataset to use\n`train-sizes` vector of double from 0 to 1, controlling the sizes of the training data.\n`options`\n   `k` At each step a k cross-validation is done\n   `metric-fn` the metric to use for evaluation the model\n   `loss-or-accuracy`   If the metric-fn calculates :loss or :accuracy\n\n", :type :var})} {:name scicloj.metamorph.ml.loss, :publics ({:name auc, :file "scicloj/metamorph/ml/loss.clj", :line 74, :arglists ([predictions labels]), :doc "Calculates area under the ROC curve. Uses AUC formula from R's 'mlr' package.\n(sum(r[i]) - n.pos * (n.pos + 1) / 2) / (n.pos * n.neg)\nSee https://github.com/mlr-org/mlr/blob/main/R/measures.R", :type :var} {:name classification-accuracy, :file "scicloj/metamorph/ml/loss.clj", :line 55, :arglists ([lhs rhs]), :doc "correct/total.\nModel output is a sequence of probability distributions.\nlabel-seq is a sequence of values.  The answer is considered correct\nif the key highest probability in the model output entry matches\nthat label.", :type :var} {:name classification-loss, :file "scicloj/metamorph/ml/loss.clj", :line 67, :arglists ([lhs rhs]), :doc "1.0 - classification-accuracy.\n", :type :var} {:name mae, :file "scicloj/metamorph/ml/loss.clj", :line 27, :arglists ([predictions labels]), :doc "mean absolute error\n", :type :var} {:name mse, :file "scicloj/metamorph/ml/loss.clj", :line 8, :arglists ([predictions labels]), :doc "mean squared error\n", :type :var} {:name rmse, :file "scicloj/metamorph/ml/loss.clj", :line 20, :arglists ([predictions labels]), :doc "root mean squared error\n", :type :var}), :doc "Simple loss functions.\n"} {:name scicloj.metamorph.ml.malli, :publics ({:name instrument-mm, :file "scicloj/metamorph/ml/malli.clj", :line 8, :arglists ([fn]), :type :var} {:name instrument-ns, :file "scicloj/metamorph/ml/malli.clj", :line 19, :arglists ([ns]), :type :var})} {:name scicloj.metamorph.ml.metrics, :publics ({:name accuracy, :file "scicloj/metamorph/ml/metrics.clj", :line 26, :arglists ([y y_hat]), :doc "First argument `y` is the true class, `y_hat` is the predicted value.\nReturns the percentage correct.", :type :var} {:name AIC, :file "scicloj/metamorph/ml/metrics.clj", :line 268, :arglists ([model ds prediction-ds]), :type :var} {:name all-metrics, :file "scicloj/metamorph/ml/metrics.clj", :line 245, :arglists ([labels predictions label->class-fn iou-fn iou-threshold]), :doc "Returns global and per-class metrics for a given set of labels and predictions.\n- label->class-fn should take a label or prediction and return the class as a string or keyword.\n- iou-fn should take a label and prediction and return the intersection over union score\n- iou-threshold determines what iou value constitutes a matching bounding box.\n** NOTE: If labels and predictions are produced from a sequence of images,\n   ensure that the bounding boxes are shifted in each image so that there is not an overlap.", :type :var} {:name BIC, :file "scicloj/metamorph/ml/metrics.clj", :line 282, :arglists ([model ds prediction-ds]), :type :var} {:name eer-accuracy, :file "scicloj/metamorph/ml/metrics.clj", :line 150, :arglists ([y y_est] [y y_est bins]), :doc "Returns the accuracy where TPR and FPR are balanced, as well as the\nthreshold value where this balance is obtained. ROC-EER is the standard\naccuracy measurement in facial recognition.", :type :var} {:name equal-error-point, :file "scicloj/metamorph/ml/metrics.clj", :line 136, :arglists ([y y_est] [y y_est bins]), :doc "Given y and the continuous, normalized output of a predictor's estimates of\nbinary class predictions corresponding to y_hat, select the threshold which\nminimizes the difference between true and false positive rates.", :type :var} {:name error-rate, :file "scicloj/metamorph/ml/metrics.clj", :line 17, :arglists ([y y_hat]), :doc "First argument `y` is the true class, `y_hat` is the predicted value.\nReturns the percentage error rate.", :type :var} {:name false-negatives, :file "scicloj/metamorph/ml/metrics.clj", :line 39, :arglists ([y y_hat]), :doc "Returns array with 1. values assigned to false negatives.\n", :type :var} {:name false-positives, :file "scicloj/metamorph/ml/metrics.clj", :line 33, :arglists ([y y_hat]), :doc "Returns array with 1. values assigned to false positives.\n", :type :var} {:name fnr, :file "scicloj/metamorph/ml/metrics.clj", :line 87, :arglists ([y y_hat]), :doc "The false negative rate, using the strict ROC definition.\n", :type :var} {:name fpr, :file "scicloj/metamorph/ml/metrics.clj", :line 73, :arglists ([y y_hat]), :doc "The false negative rate, using the strict ROC definition.\n", :type :var} {:name precision, :file "scicloj/metamorph/ml/metrics.clj", :line 65, :arglists ([y y_hat]), :doc "Returns precision for a binary classifier, a measure of false positive rate\n", :type :var} {:name recall, :file "scicloj/metamorph/ml/metrics.clj", :line 57, :arglists ([y y_hat]), :doc "Returns recall for a binary classifier, a measure of false negative rate\n", :type :var} {:name roc-curve, :file "scicloj/metamorph/ml/metrics.clj", :line 117, :arglists ([y y_est] [y y_est bins]), :doc "Compute an ROC curve with `bins` level of discretization for threshold values\nbetween 0.0 and 1.0 to compute true and false positive rates for.\n\nThis is not at all an ideal implementation, just a stand in that is useful\nfor certain problems until a real alternative is provided.", :type :var} {:name threshold, :file "scicloj/metamorph/ml/metrics.clj", :line 93, :arglists ([y_est thresh]), :doc "Return a binary mask of all values above threshold.\n", :type :var} {:name tpr, :file "scicloj/metamorph/ml/metrics.clj", :line 80, :arglists ([y y_hat]), :doc "The true positive rate, using the strict ROC definition.\n", :type :var} {:name true-negatives, :file "scicloj/metamorph/ml/metrics.clj", :line 51, :arglists ([y y_hat]), :doc "Returns array with 1. values assigned to true negatives.\n", :type :var} {:name true-positives, :file "scicloj/metamorph/ml/metrics.clj", :line 45, :arglists ([y y_hat]), :doc "Returns array with 1. values assigned to true positives.\n", :type :var} {:name unit-space, :file "scicloj/metamorph/ml/metrics.clj", :line 98, :arglists ([divs]), :doc "Returns an array with divs+1 values that evenly divide a space from 0.0 to\n1.0, inclusive.", :type :var} {:name wrongs, :file "scicloj/metamorph/ml/metrics.clj", :line 10, :arglists ([y y_hat]), :doc "Given `y` array of ground truth labels and `y_hat` classifier predictions,\nreturns array with 1.0 values where `y` does not equal `y_hat`.", :type :var}), :doc "Excellent metrics tools from the cortex project.\n"} {:name scicloj.metamorph.ml.preprocessing, :publics ({:name min-max-scale, :file "scicloj/metamorph/ml/preprocessing.clj", :line 52, :arglists ([columns-selector {:keys [min max], :or {min -0.5, max 0.5}, :as options}]), :doc "Metamorph transfomer, which scales the column data into a given range.\n\n`columns-selector` tablecloth columns-selector to choose columns to work on\n`meta-field` tablecloth meta-field working with `columns-selector`\n\n`options` Options for scaler, can take:\n    `min` Minimal value to scale to (default -0.5)\n    `max` Maximum value to scale to (default 0.5)\n\nmetamorph                            | .\n-------------------------------------|----------------------------------------------------------------------------\nBehaviour in mode :fit               | Scales the dataset at key `:metamorph/data` and stores the trained model in ctx under key at `:metamorph/id`\nBehaviour in mode :transform         | Reads trained min-max-scale model from ctx and applies it to data in `:metamorph/data`\nReads keys from ctx                  | In mode `:transform` : Reads trained model to use for from key in `:metamorph/id`.\nWrites keys to ctx                   | In mode `:fit` : Stores trained model in key $id\n\n", :type :var} {:name std-scale, :file "scicloj/metamorph/ml/preprocessing.clj", :line 27, :arglists ([columns-selector meta-field {:keys [mean? stddev?], :or {mean? true, stddev? true}, :as options}] [columns-selector options]), :doc "Metamorph transfomer, which centers and scales the dataset per column.\n\n`columns-selector` tablecloth columns-selector to choose columns to work on\n`meta-field` tablecloth meta-field working with `columns-selector`\n\n\n`options` are the options for the scaler and can take:\n   `mean?` If true (default), the data gets shifted by the column means, so 0 centered\n   `stddev?` If true (default), the data gets scaled by the standard deviation of the column\n\nmetamorph                            | .\n-------------------------------------|----------------------------------------------------------------------------\nBehaviour in mode :fit               | Centers and scales the dataset at key `:metamorph/data` and stores the trained model in ctx under key at `:metamorph/id`\nBehaviour in mode :transform         | Reads trained std-scale model from ctx and applies it to data in `:metamorph/data`\nReads keys from ctx                  | In mode `:transform` : Reads trained model to use for from key in `:metamorph/id`.\nWrites keys to ctx                   | In mode `:fit` : Stores trained model in key $id\n\n", :type :var})} {:name scicloj.metamorph.ml.tools, :publics ({:name dissoc-in, :file "scicloj/metamorph/ml/tools.clj", :line 13, :arglists ([m ks]), :doc "Dissociate a value in a nested assocative structure, identified by a sequence\nof keys. Any collections left empty by the operation will be dissociated from\ntheir containing structures.", :type :var} {:name multi-dissoc-in, :file "scicloj/metamorph/ml/tools.clj", :line 28, :arglists ([m kss]), :type :var} {:name pp-str, :file "scicloj/metamorph/ml/tools.clj", :line 35, :arglists ([x]), :type :var})} {:name scicloj.metamorph.ml.toydata, :publics ({:name breast-cancer-ds, :file "scicloj/metamorph/ml/toydata.clj", :line 63, :arglists ([]), :type :var} {:name diabetes-ds, :file "scicloj/metamorph/ml/toydata.clj", :line 24, :arglists ([]), :type :var} {:name iris-ds, :file "scicloj/metamorph/ml/toydata.clj", :line 47, :arglists ([]), :type :var} {:name mtcars-ds, :file "scicloj/metamorph/ml/toydata.clj", :line 112, :arglists ([]), :type :var} {:name sonar-ds, :file "scicloj/metamorph/ml/toydata.clj", :line 10, :arglists ([]), :type :var} {:name titanic-ds-split, :file "scicloj/metamorph/ml/toydata.clj", :line 100, :arglists ([]), :type :var})} {:name scicloj.metamorph.ml.toydata.ggplot, :publics ({:name diamonds, :file "scicloj/metamorph/ml/toydata/ggplot.clj", :line 17, :type :var} {:name ecomonics, :file "scicloj/metamorph/ml/toydata/ggplot.clj", :line 20, :type :var} {:name ecomonics_long, :file "scicloj/metamorph/ml/toydata/ggplot.clj", :line 23, :type :var} {:name faithfuld, :file "scicloj/metamorph/ml/toydata/ggplot.clj", :line 26, :type :var} {:name luv_colours, :file "scicloj/metamorph/ml/toydata/ggplot.clj", :line 29, :type :var} {:name midwest, :file "scicloj/metamorph/ml/toydata/ggplot.clj", :line 32, :type :var} {:name mpg, :file "scicloj/metamorph/ml/toydata/ggplot.clj", :line 35, :type :var} {:name msleep, :file "scicloj/metamorph/ml/toydata/ggplot.clj", :line 38, :type :var} {:name presidential, :file "scicloj/metamorph/ml/toydata/ggplot.clj", :line 41, :type :var} {:name seals, :file "scicloj/metamorph/ml/toydata/ggplot.clj", :line 44, :type :var} {:name txhousing, :file "scicloj/metamorph/ml/toydata/ggplot.clj", :line 47, :type :var})} {:name scicloj.metamorph.ml.verify, :publics ({:name basic-classification, :file "scicloj/metamorph/ml/verify.clj", :line 55, :arglists ([options-map max-avg-loss] [options-map]), :type :var} {:name basic-regression, :file "scicloj/metamorph/ml/verify.clj", :line 36, :arglists ([options-map max-avg-loss] [options-map]), :type :var} {:name classification-titanic*, :file "scicloj/metamorph/ml/verify.clj", :line 22, :type :var} {:name regression-iris*, :file "scicloj/metamorph/ml/verify.clj", :line 16, :type :var} {:name target-colname, :file "scicloj/metamorph/ml/verify.clj", :line 13, :type :var})} {:name scicloj.metamorph.ml.viz, :publics ({:name apply-xform-kvs, :file "scicloj/metamorph/ml/viz.clj", :line 8, :arglists ([spec kvs]), :type :var} {:name confusion-matrix, :file "scicloj/metamorph/ml/viz.clj", :line 56, :arglists ([predicted-labels labels opts hanami-opts] [predicted-labels labels]), :doc "Generates a confusin matrix plot out of `predicted-labels` and `labels`\n\n`opts`\n  `normalize` : Can be :none (default) or :all and decides if the values in the matrix are counts or percentages.\n\n`hanami-opts` Options passed to hanami to control the plot. Can be any of the default hanami\nsubstituions keys.\n\n", :type :var} {:name learning-curve, :file "scicloj/metamorph/ml/viz.clj", :line 12, :arglists ([dataset pipe-fn train-sizes lc-opts hanami-opts] [dataset pipe-fn lc-opts] [dataset pipe-fn lc-opts hanami-opts]), :doc "Generates a learning  curve.\n\nThe functions splits the dataset in a fixed size test set\nand increasingly larger training sets. A model is trained at each\nstep and evaluated.\n\nReturns a vega lite spec of the learninig curve plot.\n\n`dataset` the TMD dataset to use\n`train-sizes` vector of double from 0 to 1, controlling the sizes of the training data.\n`lc-opts`\n   `k` At each step a k cross-validation is done\n   `metric-fn` the metric to use for evaluation the model\n   `loss-or-accuracy`   If the metric-fn calculates :loss or :accuracy\n`hanami-opts` Options passed to hanami to control the plot. Can be the default hanami\n substituions keys or:\n     `TRAIN-COLOR:`   Color used for the train curve (default: blue)\n     `TEST-COLOR:`    Color used for the test curve (default: orange)\n\n", :type :var})} {:name scicloj.metamorph.ml.viz.confusionmatrix, :publics ({:name cm-values, :file "scicloj/metamorph/ml/viz/confusionmatrix.clj", :line 38, :arglists ([predicted-labels labels opts]), :type :var} {:name confusion-matrix-chart, :file "scicloj/metamorph/ml/viz/confusionmatrix.clj", :line 19, :arglists ([values]), :type :var} {:name layer, :file "scicloj/metamorph/ml/viz/confusionmatrix.clj", :line 10, :type :var})} {:name scicloj.metamorph.ml.viz.learning-curve, :publics ({:name errorband-encoding-test, :file "scicloj/metamorph/ml/viz/learning_curve.clj", :line 21, :type :var} {:name errorband-encoding-train, :file "scicloj/metamorph/ml/viz/learning_curve.clj", :line 9, :type :var} {:name layer, :file "scicloj/metamorph/ml/viz/learning_curve.clj", :line 49, :type :var} {:name metric-encoding, :file "scicloj/metamorph/ml/viz/learning_curve.clj", :line 34, :type :var} {:name spec, :file "scicloj/metamorph/ml/viz/learning_curve.clj", :line 59, :arglists ([lc-vl-data]), :type :var} {:name vl-data, :file "scicloj/metamorph/ml/viz/learning_curve.clj", :line 89, :arglists ([lc-rf]), :type :var})})}, :pom-str "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<project xmlns=\"http://maven.apache.org/POM/4.0.0\" xmlns:xsi=\"http://www.w3.org/2001/XMLSchema-instance\" xsi:schemaLocation=\"http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd\">\n  <modelVersion>4.0.0</modelVersion>\n  <packaging>jar</packaging>\n  <groupId>scicloj</groupId>\n  <artifactId>metamorph.ml</artifactId>\n  <version>0.7.10</version>\n  <name>metamorph.ml</name>\n  <dependencies>\n    <dependency>\n      <groupId>org.clojure</groupId>\n      <artifactId>clojure</artifactId>\n      <version>1.11.2</version>\n    </dependency>\n    <dependency>\n      <groupId>generateme</groupId>\n      <artifactId>fastmath</artifactId>\n      <version>2.4.0</version>\n    </dependency>\n    <dependency>\n      <groupId>aerial.hanami</groupId>\n      <artifactId>aerial.hanami</artifactId>\n      <version>0.19.0</version>\n    </dependency>\n    <dependency>\n      <groupId>scicloj</groupId>\n      <artifactId>tablecloth</artifactId>\n      <version>7.021</version>\n    </dependency>\n    <dependency>\n      <groupId>pppmap</groupId>\n      <artifactId>pppmap</artifactId>\n      <version>1.0.0</version>\n    </dependency>\n    <dependency>\n      <groupId>scicloj</groupId>\n      <artifactId>metamorph</artifactId>\n      <version>0.2.4</version>\n    </dependency>\n    <dependency>\n      <groupId>metosin</groupId>\n      <artifactId>malli</artifactId>\n      <version>0.14.0</version>\n    </dependency>\n  </dependencies>\n  <build>\n    <sourceDirectory>src</sourceDirectory>\n  </build>\n  <repositories>\n    <repository>\n      <id>clojars</id>\n      <url>https://repo.clojars.org/</url>\n    </repository>\n  </repositories>\n  <description>Machine learning functions for tech.ml.dataset</description>\n  <url>https://github.com/scicloj/metamorph.ml</url>\n  <licenses>\n    <license>\n      <name>Eclipse Public License</name>\n      <url>http://www.eclipse.org/legal/epl-v10.html</url>\n    </license>\n  </licenses>\n  <developers>\n    <developer>\n      <name>Carsten Behring</name>\n    </developer>\n  </developers>\n  <scm>\n    <url>https://github.com/scicloj/metamorph.ml</url>\n    <connection>scm:git:https://github.com/scicloj/metamorph.ml.git</connection>\n    <developerConnection>scm:git:https://github.com/scicloj/metamorph.ml.git</developerConnection>\n    <tag>v0.7.10</tag>\n  </scm>\n</project>\n"}