Submit
Path:
~
/
/
proc
/
thread-self
/
root
/
opt
/
alt
/
python35
/
lib64
/
python3.5
/
site-packages
/
sklearn
/
cluster
/
__pycache__
/
File Content:
birch.cpython-35.pyc
��(X�X � @ s> d d l m Z d d l Z d d l Z d d l m Z d d l m Z d d l m Z d d l m Z m Z m Z d d l m Z d d l m Z d d l m Z m Z d d l m Z d d l m Z d d l m Z d d � Z d d � Z Gd d � d e � Z Gd d � d e � Z Gd d � d e e e � Z! d S)� )�divisionN)�sparse)�sqrt� )�euclidean_distances)�TransformerMixin� ClusterMixin� BaseEstimator)�xrange)� check_array)� row_norms�safe_sparse_dot)�check_is_fitted)�NotFittedError� )�AgglomerativeClusteringc c s� | j d } | j } | j } | j } xl t | � D]^ } t j | j d � } | | | | d } } | | | � } | | | � | | <| Vq5 Wd S)z�This little hack returns a densified row when iterating over a sparse matrix, instead of constructing a sparse matrix for every row that is expensive. r r N)�shape�indices�dataZindptrr �np�zeros) �X� n_samplesZ X_indicesZX_dataZX_indptr�i�rowZstartptrZendptrZnonzero_indices� r � /birch.py�_iterate_sparse_X s r c C s� t � } t � } t | | d | j d | j �} t | | d | j d | j �} | | _ | | _ | j r� | j d k r� | | j _ | j | _ | | _ | | _ | j | _ | j d k r� | | j _ t | j d | j d d �} | j d } t j | j � | | f � } | | g \ } } | | k } x^ t | j � D]M \ } } | | rz| j | � | j | � qG| j | � | j | � qGW| | f S)a� The node has to be split if there is no place for a new subcluster in the node. 1. Two empty nodes and two empty subclusters are initialized. 2. The pair of distant subclusters are found. 3. The properties of the empty subclusters and nodes are updated according to the nearest distance between the subclusters to the pair of distant subclusters. 4. The two nodes are set as children to the two subclusters. �is_leaf� n_featuresNZY_norm_squared�squaredTr )� _CFSubcluster�_CFNoder r �child_� prev_leaf_� next_leaf_r � centroids_� squared_norm_r r Z unravel_indexZargmax� enumerate� subclusters_�append_subcluster�update)Znode� threshold�branching_factor�new_subcluster1�new_subcluster2Z new_node1Z new_node2Zdist� n_clustersZ farthest_idxZ node1_distZ node2_distZ node1_closer�idx� subclusterr r r � _split_node( sB r3 c @ sF e Z d Z d Z d d � Z d d � Z d d � Z d d � Z d S) r" a Each node in a CFTree is called a CFNode. The CFNode can have a maximum of branching_factor number of CFSubclusters. Parameters ---------- threshold : float Threshold needed for a new subcluster to enter a CFSubcluster. branching_factor : int Maximum number of CF subclusters in each node. is_leaf : bool We need to know if the CFNode is a leaf or not, in order to retrieve the final subclusters. n_features : int The number of features. Attributes ---------- subclusters_ : array-like list of subclusters for a particular CFNode. prev_leaf_ : _CFNode prev_leaf. Useful only if is_leaf is True. next_leaf_ : _CFNode next_leaf. Useful only if is_leaf is True. the final subclusters. init_centroids_ : ndarray, shape (branching_factor + 1, n_features) manipulate ``init_centroids_`` throughout rather than centroids_ since the centroids are just a view of the ``init_centroids_`` . init_sq_norm_ : ndarray, shape (branching_factor + 1,) manipulate init_sq_norm_ throughout. similar to ``init_centroids_``. centroids_ : ndarray view of ``init_centroids_``. squared_norm_ : ndarray view of ``init_sq_norm_``. c C s~ | | _ | | _ | | _ | | _ g | _ t j | d | f � | _ t j | d � | _ g | _ d | _ d | _ d S)Nr ) r, r- r r r) r r �init_centroids_� init_sq_norm_r' r$ r% )�selfr, r- r r r r r �__init__� s z_CFNode.__init__c C s� t | j � } | j j | � | j | j | <| j | j | <| j d | d � d d � f | _ | j d | d � | _ d S)Nr ) �lenr) �append� centroid_r4 �sq_norm_r5 r& r' )r6 r2 r r r r r* � s &z_CFNode.append_subclusterc C sP | j j | � } | | j | <| j | j | <| j | j | <| j | � d S)zZRemove a subcluster from a node and update it with the split subclusters. N)r) �indexr: r4 r; r5 r* )r6 r2 r. r/ Zindr r r �update_split_subclusters� s z _CFNode.update_split_subclustersc C s� | j s | j | � d S| j } | j } t j | j | j � } | d 9} | | j 7} t j | � } | j | } | j d k r.| j j | � } | s� | j | � | j | j | j | <| j | j | j | <d St | j | | � \ } } | j | | | � t | j � | j k r'd Sd Sny | j | | j � } | rm| j | j | <| j | j | <d St | j � | j k r�| j | � d S| j | � d Sd S)z&Insert a new subcluster into the node.Fg @NTg �)r) r* r, r- r �dotr& r: r' �argminr# �insert_cf_subclusterr+ r4 r; r5 r3 r= r8 �merge_subcluster) r6 r2 r, r- Z dist_matrixZ closest_indexZclosest_subclusterZ split_childr. r/ Zmergedr r r r@ � sH z _CFNode.insert_cf_subclusterN)�__name__� __module__� __qualname__�__doc__r7 r* r= r@ r r r r r" Z s . r" c @ sO e Z d Z d Z d d d � Z d d � Z d d � Z e d d � � Z d S) r! a7 Each subcluster in a CFNode is called a CFSubcluster. A CFSubcluster can have a CFNode has its child. Parameters ---------- linear_sum : ndarray, shape (n_features,), optional Sample. This is kept optional to allow initialization of empty subclusters. Attributes ---------- n_samples_ : int Number of samples that belong to each subcluster. linear_sum_ : ndarray Linear sum of all the samples in a subcluster. Prevents holding all sample data in memory. squared_sum_ : float Sum of the squared l2 norms of all samples belonging to a subcluster. centroid_ : ndarray Centroid of the subcluster. Prevent recomputing of centroids when ``CFNode.centroids_`` is called. child_ : _CFNode Child Node of the subcluster. Once a given _CFNode is set as the child of the _CFNode, it is set to ``self.child_``. sq_norm_ : ndarray Squared norm of the subcluster. Used to prevent recomputing when pairwise minimum distances are computed. Nc C sr | d k r* d | _ d | _ d | _ n; d | _ | | _ | _ t j | j | j � | _ | _ d | _ d S)Nr g r )� n_samples_� squared_sum_� linear_sum_r: r r> r; r# )r6 � linear_sumr r r r7 s z_CFSubcluster.__init__c C sh | j | j 7_ | j | j 7_ | j | j 7_ | j | j | _ t j | j | j � | _ d S)N)rF rH rG r: r r> r; )r6 r2 r r r r+ " s z_CFSubcluster.updatec C s� | j | j } | j | j } | j | j } d | | } t j | | � } d | | } | | | | } | | d k r� | | | | | f \ | _ | _ | _ | _ | _ d Sd S)zUCheck if a cluster is worthy enough to be merged. If yes then merge. r r TF���)rG rH rF r r> r: r; ) r6 Znominee_clusterr, Znew_ssZnew_lsZnew_nZ new_centroidZnew_norm� dot_productZ sq_radiusr r r rA ) s 3z _CFSubcluster.merge_subclusterc C s; d t j | j | j � } t | j | | j | j � S)z Return radius of the subclusterr rJ )r r> rH r: r rG rF r; )r6 rK r r r �radius; s z_CFSubcluster.radius) rB rC rD rE r7 r+ rA �propertyrL r r r r r! � s " r! c @ s� e Z d Z d Z d d d d d d d � Z d d d � Z d d � Z d d � Z d d d d � Z d d � Z d d � Z d d d � Z d d d � Z d S)�Bircha� Implements the Birch clustering algorithm. Every new sample is inserted into the root of the Clustering Feature Tree. It is then clubbed together with the subcluster that has the centroid closest to the new sample. This is done recursively till it ends up at the subcluster of the leaf of the tree has the closest centroid. Read more in the :ref:`User Guide <birch>`. Parameters ---------- threshold : float, default 0.5 The radius of the subcluster obtained by merging a new sample and the closest subcluster should be lesser than the threshold. Otherwise a new subcluster is started. branching_factor : int, default 50 Maximum number of CF subclusters in each node. If a new samples enters such that the number of subclusters exceed the branching_factor then the node has to be split. The corresponding parent also has to be split and if the number of subclusters in the parent is greater than the branching factor, then it has to be split recursively. n_clusters : int, instance of sklearn.cluster model, default 3 Number of clusters after the final clustering step, which treats the subclusters from the leaves as new samples. If None, this final clustering step is not performed and the subclusters are returned as they are. If a model is provided, the model is fit treating the subclusters as new samples and the initial data is mapped to the label of the closest subcluster. If an int is provided, the model fit is AgglomerativeClustering with n_clusters set to the int. compute_labels : bool, default True Whether or not to compute labels for each fit. copy : bool, default True Whether or not to make a copy of the given data. If set to False, the initial data will be overwritten. Attributes ---------- root_ : _CFNode Root of the CFTree. dummy_leaf_ : _CFNode Start pointer to all the leaves. subcluster_centers_ : ndarray, Centroids of all subclusters read directly from the leaves. subcluster_labels_ : ndarray, Labels assigned to the centroids of the subclusters after they are clustered globally. labels_ : ndarray, shape (n_samples,) Array of labels assigned to the input data. if partial_fit is used instead of fit, they are assigned to the last batch of data. Examples -------- >>> from sklearn.cluster import Birch >>> X = [[0, 1], [0.3, 1], [-0.3, 1], [0, -1], [0.3, -1], [-0.3, -1]] >>> brc = Birch(branching_factor=50, n_clusters=None, threshold=0.5, ... compute_labels=True) >>> brc.fit(X) Birch(branching_factor=50, compute_labels=True, copy=True, n_clusters=None, threshold=0.5) >>> brc.predict(X) array([0, 0, 0, 1, 1, 1]) References ---------- * Tian Zhang, Raghu Ramakrishnan, Maron Livny BIRCH: An efficient data clustering method for large databases. http://www.cs.sfu.ca/CourseCentral/459/han/papers/zhang96.pdf * Roberto Perdisci JBirch - Java implementation of BIRCH clustering algorithm https://code.google.com/archive/p/jbirch g �?�2 � Tc C s1 | | _ | | _ | | _ | | _ | | _ d S)N)r, r- r0 �compute_labels�copy)r6 r, r- r0 rQ rR r r r r7 � s zBirch.__init__Nc C s d \ | _ | _ | j | � S)z� Build a CF Tree for the input data. Parameters ---------- X : {array-like, sparse matrix}, shape (n_samples, n_features) Input data. TF)TF)�fit_� partial_fit_�_fit)r6 r �yr r r �fit� s z Birch.fitc C s� t | d d d | j �} | j } | j } | d k rE t d � � | j \ } } t | d � } t | d d � } t | d � s� | r� | r� t | | d d d | �| _ t | | d d d | �| _ | j | j _ | j | j _ t j | � st } n t } x� | | � D]� } t d | � } | j j | � } | rt | j | | � \ } } | ` t | | d d d | �| _ | j j | � | j j | � qWt j d d � | j � D� � } | | _ | j | � | S)N� accept_sparse�csrrR r z,Branching_factor should be greater than one.rT �root_rS r Tr rI Fc S s g | ] } | j � q Sr )r& )�.0Zleafr r r � <listcomp>� s z Birch._fit.<locals>.<listcomp>)r rR r, r- � ValueErrorr �getattrr" rZ � dummy_leaf_r% r$ r Zissparse�iterr r! r@ r3 r* r Z concatenate� _get_leaves�subcluster_centers_�_global_clustering)r6 r r, r- r r � partial_fitZhas_rootZ iter_func�sampler2 �splitr. r/ � centroidsr r r rU � sF z Birch._fitc C s? | j j } g } x& | d k r: | j | � | j } q W| S)z� Retrieve the leaves of the CF Node. Returns ------- leaves: array-like List of the leaf nodes. N)r_ r% r9 )r6 Zleaf_ptrZleavesr r r ra � s zBirch._get_leavesc C sJ d \ | _ | _ | d k r, | j � | S| j | � | j | � Sd S)a% Online learning. Prevents rebuilding of CFTree from scratch. Parameters ---------- X : {array-like, sparse matrix}, shape (n_samples, n_features), None Input data. If X is not provided, only the global clustering step is done. TFN)TF)rT rS rc � _check_fitrU )r6 r rV r r r rd � s zBirch.partial_fitc C si t | d � } t | d � } | p' | s6 t d � � | re | j d | j j d k re t d � � d S)Nrb rT z#Fit training data before predictingr zETraining data and predicted data do not have same number of features.)�hasattrr r rb r] )r6 r Z is_fittedZhas_partial_fitr r r rh s #zBirch._check_fitc C se t | d d �} | j | � t | | j j � } | d 9} | | j 7} | j t j | d d �S)ae Predict data using the ``centroids_`` of subclusters. Avoid computation of the row norms of X. Parameters ---------- X : {array-like, sparse matrix}, shape (n_samples, n_features) Input data. Returns ------- labels: ndarray, shape(n_samples) Labelled data. rX rY r Zaxisr rJ ) r rh r rb �T�_subcluster_norms�subcluster_labels_r r? )r6 r Zreduced_distancer r r �predict s z Birch.predictc C s t | d � t | | j � S)a� Transform X into subcluster centroids dimension. Each dimension represents the distance from the sample point to each cluster centroid. Parameters ---------- X : {array-like, sparse matrix}, shape (n_samples, n_features) Input data. Returns ------- X_trans : {array-like, sparse matrix}, shape (n_samples, n_clusters) Transformed data. rb )r r rb )r6 r rV r r r � transform' s zBirch.transformc C s0 | j } | j } | d k o$ | j } d } t | t � rl t d | j � } t | � | j k r� d } n( | d k r� t | d � r� t d � � t | j d d �| _ | d k s� | r� t j t | � � | _ | rt j d t | � | j f � n | j | j � | _ | r,| j | � | _ d S) zN Global clustering for the subclusters obtained after fitting NFr0 T� fit_predictz:n_clusters should be an instance of ClusterMixin or an intr zTNumber of subclusters found (%d) by Birch is less than (%d). Decrease the threshold.)r0 rb rQ � isinstance�intr r8 ri r] r rk r Zarangerl �warnings�warnro rm Zlabels_)r6 r Z clustererrg rQ Znot_enough_centroidsr r r rc ; s0 zBirch._global_clustering) rB rC rD rE r7 rW rU ra rd rh rm rn rc r r r r rN D s Q 3 rN )"Z __future__r rr Znumpyr Zscipyr �mathr Zmetrics.pairwiser �baser r r Zexternals.six.movesr Zutilsr Z utils.extmathr r Zutils.validationr � exceptionsr Z hierarchicalr r r3 �objectr" r! rN r r r r �<module> s" 2�Q
Edit
Rename
Chmod
Delete
FILE
FOLDER
Name
Size
Permission
Action
__init__.cpython-35.opt-1.pyc
1160 bytes
0644
__init__.cpython-35.pyc
1160 bytes
0644
_feature_agglomeration.cpython-35.opt-1.pyc
2478 bytes
0644
_feature_agglomeration.cpython-35.pyc
2478 bytes
0644
affinity_propagation_.cpython-35.opt-1.pyc
9332 bytes
0644
affinity_propagation_.cpython-35.pyc
9332 bytes
0644
bicluster.cpython-35.opt-1.pyc
17924 bytes
0644
bicluster.cpython-35.pyc
17924 bytes
0644
birch.cpython-35.opt-1.pyc
17712 bytes
0644
birch.cpython-35.pyc
17712 bytes
0644
dbscan_.cpython-35.opt-1.pyc
11471 bytes
0644
dbscan_.cpython-35.pyc
11471 bytes
0644
hierarchical.cpython-35.opt-1.pyc
26136 bytes
0644
hierarchical.cpython-35.pyc
26172 bytes
0644
k_means_.cpython-35.opt-1.pyc
43828 bytes
0644
k_means_.cpython-35.pyc
43903 bytes
0644
mean_shift_.cpython-35.opt-1.pyc
14364 bytes
0644
mean_shift_.cpython-35.pyc
14364 bytes
0644
setup.cpython-35.opt-1.pyc
1312 bytes
0644
setup.cpython-35.pyc
1312 bytes
0644
spectral.cpython-35.opt-1.pyc
15802 bytes
0644
spectral.cpython-35.pyc
15802 bytes
0644
N4ST4R_ID | Naxtarrr
