refactor: manually extracted refined types (#46)

* Refactor the manually extracted refined types for classes KMeans, PCA, TruncatedSVD, CountVectorizer and TfIdfVectorizer

* refactor refinedtypes for DecisionTreeClassifier

* refactor json for model_selection,neighbors,preprocessing,svm,tree

* refactoring for metrics,linearmodel,impute

* refactored json for feature_extraction

* updated BoundaryTypes and NamedTypes and the remaining classes

* Refactor manually extracted refined types for cluster, decomposition, ensemble

* Refactor manually extracted refined types for feature_extraction, impute and linear_model

* refactored manually extracted refined types for metrics,model_selection,neighbors,preprocessing,svm,tree

* Format the json files from the manual extraction of refined types

* Extract the boundary types in another json and the docstring part that is related to the boundary

* style: apply automatic fixes of linters

Co-authored-by: Dushko Klincharov <klincarov@halicea.com>
Co-authored-by: mmdoja <mmdoja@gmail.com>
Co-authored-by: duklin <duklin@users.noreply.github.com>

Author: 3 people

refined_types/CategorizedBoundaries.md

@@ -0,0 +1,16 @@
+Keywords for boundaries extraction:
+- `non-negative float`
+- `Must be of range [0.0, infinity)`
+- `Must be between 0 and 1`
+- `Thus, max_samples should be in the interval (0.0, 1.0]`
+- `float > 0 and <= 1`
+- `If float in range [0.0, 1.0]`
+- `float in range [0.0, 1.0]`
+- `Must be strictly positive`
+- `tuple (q_min, q_max), 0.0 < q_min < q_max < 100.0`
+- `The ElasticNet mixing parameter, with 0 <= l1_ratio <= 1`
+- `must be a positive float`
+- `For the liblinear and lbfgs solvers set verbose to any positive number for verbosity`
+- `Must be at least 2`
+- `If float, should be between 0.0 and 1.0 and represent the proportion of the dataset to include in the test split`
+- `If float, should be between 0.0 and 1.0 and represent the proportion of the dataset to include in the train split`