Fix pipeline clone

.github/workflows/test.yml

@@ -33,12 +33,9 @@ jobs:
 
     - name: Test with pytest
       run: |
-        coverage run --source=. -m pytest
+        coverage run --source=. -m pytest --nbmake
         coverage report -m
 
-    - name: Test of jupyter notebook
-      run: |
-        pytest --nbmake
 
     - name: Coveralls
       uses: AndreMiras/coveralls-python-action@develop

ceruleo/dataset/ts_dataset.py

@@ -1,6 +1,7 @@
 from collections.abc import Iterable
 from pathlib import Path
 from re import S
+from types import EllipsisType
 from typing import Any, List, Optional, Tuple, Union
 
 import numpy as np
@@ -101,6 +102,7 @@ def __call__(self, i):
     def get_features_of_life(self, i: int) -> pd.DataFrame:
         return self[i]
 
+
     def __getitem__(
         self, i: Union[int, Iterable]
     ) -> Union[pd.DataFrame, "FoldedDataset"]:
@@ -126,18 +128,24 @@ def __getitem__(
             )
         if TENSORFLOW_ENABLED and isinstance(i, tf.Tensor):
             return self.get_time_series(i.ref())
+
+
+
         if isinstance(i, Iterable):
             if not all(isinstance(item, (int, np.integer)) for item in i):
-                raise ValueError("Invalid iterable index passed")
+                if len(i) == 2:       
+                    if not isinstance(i[1], EllipsisType):
+                        raise ValueError("Invalid iterable index passed")
+                    i = i[0]        
 
             return FoldedDataset(self, i)
         else:
             df = self.get_time_series(i)
             return df
 
     @property
-    def shape(self) -> Tuple[int, int]:
-        return (self.n_time_series, 1)
+    def shape(self) -> Tuple[int]:
+        return (self.n_time_series,)
 
     def __len__(self) -> int:
         """

ceruleo/transformation/features/scalers.py

@@ -183,8 +183,10 @@ def transform(self, X: pd.DataFrame) -> pd.DataFrame:
         try:
             divisor = self.data_max - self.data_min
 
+
+
             mask = np.abs((divisor)) > 1e-25
-            X = X.copy()
+            X = X.astype(float)
             X.loc[:, mask] = (
                 (X.loc[:, mask] - self.data_min[mask])
                 / divisor[mask]

ceruleo/transformation/features/selection.py

@@ -52,6 +52,7 @@ def fit(self, df:pd.DataFrame, y=None):
             features = list(set(df.columns))
         self.features_computed_ = sorted(features)
         return self
+        return X.loc[:, self.features_computed_].copy()
 
     def transform(self, X:pd.DataFrame) -> pd.DataFrame:
         """ 

ceruleo/transformation/functional/mixin.py

@@ -1,12 +1,17 @@
-import hashlib
 import uuid
+from copy import deepcopy
 from typing import List, Optional, Union
 
-from numpy.lib.arraysetops import isin
 from sklearn.base import BaseEstimator
 
 
 class TransformerStepMixin(BaseEstimator):
+    name: Optional[str]
+    previous: List["TransformerStepMixin"]
+    next: List["TransformerStepMixin"]
+    uuid: str
+    prefer_partial_fit: bool
+
     def __init__(self, *, name: Optional[str] = None, prefer_partial_fit:bool =True):
         self.name_ = name
         self.previous = []
@@ -65,3 +70,8 @@ def description(self):
 
     def __str__(self):
         return self.name
+
+
+    def clear_connections(self):
+        self.previous = []
+        self.next = []

ceruleo/transformation/functional/pipeline/pipeline.py

@@ -1,10 +1,6 @@
-import shelve
-import uuid
-from pathlib import Path
-from typing import Iterable, List, Optional, Tuple, Union
+from typing import Iterable, List, Union
 
 import pandas as pd
-from ceruleo import CACHE_PATH
 from ceruleo.dataset.ts_dataset import AbstractPDMDataset
 from ceruleo.transformation.functional.graph_utils import (
     dfs_iterator,
@@ -14,7 +10,7 @@
 from ceruleo.transformation.functional.pipeline.runner import CachedPipelineRunner
 from ceruleo.transformation.functional.transformerstep import TransformerStep
 from sklearn.base import BaseEstimator, TransformerMixin
-
+from sklearn.base import clone as sklearn_clone
 
 class Pipeline(BaseEstimator, TransformerMixin):
     """Transformation pipeline
@@ -24,6 +20,11 @@ class Pipeline(BaseEstimator, TransformerMixin):
         cache_type: Cache storage mode
     """
 
+    final_step: TransformerStep
+    fitted_: bool
+    cache_type: CacheStoreType
+    runner: CachedPipelineRunner
+
     def __init__(self, final_step, cache_type: CacheStoreType = CacheStoreType.MEMORY):
         self.final_step = final_step
         self.fitted_ = False
@@ -115,6 +116,19 @@ def get_params(self, deep: bool = False):
                 for k in p.keys():
                     params[f"{node.name}__{k}"] = p[k]
         return params
+
+    def __sklearn_clone__(self):
+        g = {node: sklearn_clone(node) for node  in dfs_iterator(self.final_step)}
+        for k in g.keys():
+            g[k].clear_connections()
+        for node in dfs_iterator(self.final_step):
+            for next_node in node.next:
+                g[node].add_next(g[next_node])
+
+        return Pipeline(
+            final_step=g[self.final_step],
+            cache_type=self.cache_type
+        )
 
 
 def make_pipeline(

docs/dataset/analysis/Sensor Validation.ipynb

@@ -99,15 +99,15 @@
    "outputs": [],
    "source": [
     "from ceruleo.dataset.analysis.numerical_features import analysis\n",
-    "from ceruleo.transformation.functional.transformers import Transformer\n",
-    "from ceruleo.transformation.features.selection import ByNameFeatureSelector, ByTypeFeatureSelector\n",
     "from ceruleo.iterators.iterators import RelativeToEnd\n",
-    "from ceruleo.transformation.features.slicing import SliceRows\n",
-    "from ceruleo.transformation.functional.pipeline.pipeline import make_pipeline\n",
     "from ceruleo.transformation.features.resamplers import IndexMeanResampler\n",
-    "from ceruleo.transformation.features.transformation import Clip\n",
+    "from ceruleo.transformation.features.selection import (\n",
+    "    ByNameFeatureSelector,\n",
+    ")\n",
     "from ceruleo.transformation.features.slicing import SliceRows\n",
-    "from ceruleo.iterators.iterators import RelativeToEnd"
+    "from ceruleo.transformation.features.transformation import Clip\n",
+    "from ceruleo.transformation.functional.pipeline.pipeline import make_pipeline\n",
+    "from ceruleo.transformation.functional.transformers import Transformer"
    ]
   },
   {