As asked by @jacobmontiel, here is the first implementation of delayed evaluation. It is already stable, but we still need to address some features, like dealing with a delay by a number of samples and adding a delay to all the stream. We need to check if there are ways to improve performance by removing pandas data frame from TimeManager too. Finally, we need to create a dataset to test this new feature.

Changes proposed in this pull request:

• Created TemporalDataStream, to handle temporal data in a stream.
• Created EvaluatePrequentialDelayed,
• Created TimeManager, to handle delay queue even outside evaluation

Known bugs:

• Sometimes progress bar is printed a lot of times when reaching 100%, and sometimes it reaches more than 100%. I tried to fix that in update_metrics method, but nothing fixed it yet.

Checklist

• [x] Code complies with PEP-8 and is consistent with the framework.
• [x] Code is properly documented.
• [x] Tests are included for new functionality or updated accordingly.
• [x] Travis CI build passes with no errors.
• [x] Test Coverage is maintained (threshold is -0.2%).
• [x] Files changed (update, add, delete) are in the PR's scope (no extra files are included).

Fixes # .

Changes proposed in this pull request: Add LearnNSE algorithm Compare to MOA: SEAGenerator with window size of 1000: MOA mean acc: 83.38 multiflow learnNSE: 95.58 RandomTreeGenerator with window size of 1000: MOA mean acc: 68.12 multiflow learnNSE: 63.17

The only difference is I use GaussianNaiveBayes and I am not sure which Naive Bayes MOA use

@scikit-multiflow/scikit-multiflow-repo-admins

Checklist

• [x] Code complies with PEP-8 and is consistent with the framework.
• [x] Code is properly documented.
• [x] Tests are included for new functionality.
• [x] Travis CI build passes.
• [x] Test Coverage is maintained (threshold 0.2%)

Changes

Fixes #62, #63, #65.

Changes proposed in this pull request:

• Implementation of predict_proba for Adaptive Random Forest (#62).
• Fix a bug where weights were incorrectly passed (#63).
• Above bugfix solves #65 (technically only tested j=2,5,10 and rest still running).

Remarks

In the discussion of #63 it was mentioned that the fix reduces performance. However this was based on just the single unit test. Using the same unit test, but changing the test set (by increasing the initial value of cnt), we see the following values:

| cnt | master | PR | |---|---|---| | 0 | 0.6531 | 0.4694 | | 1 | 0.5306 | 0.4694 | | 2 | 0.4082 | 0.4694 | | 3 | 0.5310 | 0.5306 | | 4 | 0.3877 | 0.5510 | | 5 | 0.5306 | 5714 | | 6 | 0.5306 | 0.5918 | | 7 | 0.4287 | 0.4898 | | 8 | 0.408 | 0.5510 | | 9 | 0.5510 | 0.5918 |

While we can do more thorough evaluations, for me it seems that the initial set-up just was an outlier, and there is no real performance decrease. Further motivation is that if I increase the training set size on the PR version, performance does increase. On the PR branch, using 150 samples for training results in a performance of 0.4694 (with cnt=0 determining the test set). Increasing this to 1500 increases performance to 0.5714, increasing the training set again to 15000 results in a performance of 0.7142.

Checklist

• [x] Code complies with PEP-8 and is consistent with the framework.
• [x] Code is properly documented.
• [x] Tests are included for new functionality.
• [x] Travis CI build passes.
• [x] Test Coverage is maintained (threshold 0.2%)

@scikit-multiflow/scikit-multiflow-repo-admins

why this error happened in evaluation for some csv file when filestream ? hi , when i want to Filestream from a csv file i got this message error ."Exception has occurred: TypeError 'NoneType' object is not subscriptable." for example for music.csv in datasets folder, i got this message too. for some datasets it works fine , but for some doesnt. for ex, i generate a datastream and save it to csv file and when i want to filestream it , i got that error message.for others like moving_squares or sea_stream it works. i want to generate a stream from stream generate and save the stream generator to a csv file and add come columns to that csv file and then Filestream that updated csv file , but i got this error in evaluation. ,first i think it was for float numbers , i cant solve the problem . for some csv file int values it does works .

@jacobmontiel

[Note: For comments and usage-related questions please consider the users group.

This is just a question. Suppose my samples are produced as a stream, but that I cannot know the response before a certain delay (during which new samples are produced and need to be predicted). In the simple examples in the doc, it looks like the model is evaluated right after it made its prediction. Is there a way to somehow delay the evaluation, allowing for other samples to be predicted in the mean time?

Streaming Half-Space-Trees (HS-Trees), a fast one-class anomaly detector for evolving data streams.

Theory and the pseudocode algorithm can be reviewed in: S. C. Tan, K. M. Ting, and T. F. Liu, “Fast anomaly detection for streaming data,” in IJCAI Proceedings-International Joint Conference on Artificial Intelligence, 2011, vol. 22, no. 1, pp. 1511–1516.:

Changes proposed in this pull request:

• An implementation of streaming Half Space Trees.
• [TODO] Test is missing on this commit. I have a dummy branch where I created some tests, but it is reading from a big csv file which was mentioned in the paper. If you have any suggestion on how to integrate any tests on this PR I would be happy. Link for the dummy brach for testing: https://github.com/mertozer94/scikit-multiflow/tree/Silly-tests-Half-Space-Trees

Checklist

• [x] Code complies with PEP-8 and is consistent with the framework.
• [x] Code is properly documented.
• [x] Tests are included for new functionality or updated accordingly.
• [x] Travis CI build passes with no errors.
• [ ] Test Coverage is maintained (threshold is -0.2%).
• [x] Files changed (update, add, delete) are in the PR's scope (no extra files are included).

Hi @scikit-multiflow/scikit-multiflow-repo-admins,

This PR is focused on issues #18 and #56.

I've created a RunningTimeMeasurements class which accounts for the training, testing and total running time of each evaluated predictor. To ignore the time spent on other operations I had to call the running accounting methods directly before the partial_fit and predict methods on evaluate_prequential and evaluate_holdout. For instance:

https://github.com/smastelini/scikit-multiflow/blob/05f07b67809807d2564f7ad04faa35b55d46f85c/src/skmultiflow/evaluation/evaluate_prequential.py#L277-L279

Do you have any other suggestions?

Regarding the model size accounting, a function was added in skmultiflow/utils/utils.py. This tool is only called when the metrics are updated in the skmultiflow/evaluation/base_evaluator.py. The default returning unit is MB.

In addition, I provide tests for the new functionalities.

The visualizations for the new metrics were implemented as subplots without axes ticks and bounding boxes, where the measurements are annotated. That was the best solution I could find, considering the auto re-scaling of the elements that is performed periodically. Could a better option be implemented? I would happily implement the required changes.

In summary, the changes proposed in this pull request are:

• Running time is added as a new performance metric
• Model size is added as a new performance metric
• Automated tests and visualizations for the new features are also added

Checklist:

• [x] Code complies with PEP-8 and is consistent with the framework.
• [x] Code is properly documented.
• [x] Tests are included for new functionality.
• [x] Travis CI build passes.
• [ ] Test Coverage is maintained (threshold 0.2%)

• An observation: Some of the current existing commits in skmultiflow/MASTER might appear as authored by @jacobmontiel and commited by me in this PR. This was an effect of the "git rebase" I performed in my commits to avoid polluting the timeline. I merged many minor incremental changes into few steps. Sorry for the possible mess...

Thanks in advance for the help!

This PR addresses Issue #148.

Due to the large number of necessary changes in the PR, I would like to request some help to verify and discuss the validity of the performed operations.

The main idea in this PR is to increase modularity and the organization in the skmultiflow.trees modules. To this end, I envisioned the tree implementations as five main elements:

• The tree core codes, per se
• Their nodes (split nodes, prediction nodes, and so forth)
• Their attribute observers (which monitor the attribute/target combinations to enable split attempts)
• The chosen split criterion to growth
• The types of attribute tests performed (binary tests, for instance)

As a result, I've created subfolders for the last 4 aspects, respectively, nodes, attribute_observer, split_criterion, and attribute_test. These names are just a sketch. Any suggestions for a better nomenclature are welcomed!

The needed changes for attribute_observer, split_criterion, and attribute_test were pretty straightforward. I had just to update some imports and configure/create some __init__.py files. The changes for the leaf predictors, on the other hand, caused more impact. To decouple the inner classes from the trees, I had to create a lot of new files (one for each kind of leaf predictor, plus some basic node structures from HoeffdingTree).

As I mentioned in Issue #148, I believe that there might be some redundant code among the plethora of leaf node classes. This might be a task for future improvements in the framework. Additionally, I believe that such refactoring of the tree module could help with Issue #138. Maybe we could extend the renaming process to the leaf nodes as well. I've tried to standardize the names of the new files as much as I could. I had to rename some classes to comply with the name of the trees they belong to.

Nevertheless, there might be still some inconsistencies in the names. Unlike the tree algorithms (Issue #138), perhaps we could standardize the leaves names in this PR, since the users do not access them directly via the API. There would be no necessity to issue deprecation warnings in this case (IMHO :-D). What do you think, @jacobmontiel? (On the other hand, the contributors would have to adapt to the new changes)

I also updated the documentation to reflect the changes. I noticed that some of the leaf nodes do not have (a complete) documentation. With the refactoring process, this became more apparent. We could either add the missing documentation in this PR or create some issues about this need. Any opinions on that matter?

I also edited some of the code to comply with PEP8. The changes were mostly some line breaks for very long lines.

As much as all the tests are passing (at least in my local machine), given the extent of the changes, I would appreciate some help in verifying possible missing/incorrect imports. It is also possible that there are unnecessary imports I may not have noticed.

It is worth noting that we could apply the same refactoring to the skmultiflow.rules module.

To finalize, I would like to thank you @jacobmontiel, @hmgomes and the skmultiflow community in advance for the help and patience during this PR.

Any comments and/or requests for changes are welcomed. I believe that the code at its current state is the first step to achieve the desired goals for this PR.

Cheers, and sorry for the long description. Saulo

Changes proposed in this pull request:

• To streamline the Hoeffding Tree algorithms separating their leaf nodes, attribute observers, split criteria, and attribute tests.
• Improvements in some __init__.py files (adding missing exports for some classes/functions)
• Updates in the documentation
• Addresses #148.

Checklist

• [x] Code complies with PEP-8 and is consistent with the framework.
• [ ] Code is properly documented.
• [x] Tests are included for new functionality or updated accordingly.
• [x] Travis CI build passes with no errors.
• [ ] Test Coverage is maintained (threshold is -0.2%).
• [x] Files changed (update, add, delete) are in the PR's scope (no extra files are included).

This PR resumes the work (PR #145) started by Alban to implement the AdaptiveRandomForestRegressor (ARFReg). The implementation was finalized and tested. I have also added some additional options, such as different aggregation and weighting strategies, and different kinds of setups to track concept drifts.

As base learners, the Forest uses trees that derive from HoeffdingTreeRegressor (HTR), in detriment of FIMT-DD (as in MOA). HTRs do not have inner concept drift adaptation mechanisms; however, such action is performed at higher abstraction levels, by the ensemble algorithm. This action repeats in the classification version of Adaptive Random Forests.

In addition to ARFReg, this PR also streamlines the usage of attribute observers for regression. It adds a new and improved observer (derived from the multi-target regression observer) that works with both single-target and multi-target tasks. This enhanced NumericAttributeRegressionObserver also enables the support to instance weighting, an essential aspect for bagging-based algorithms. Hence, just a single observer is used for all the regression trees.

Changes proposed in this pull request:

• Finishes work initiated in PR #145
• Streamline regression observers

Checklist

Implementation

• [x] Implementation is correct (it performs its intended function).
• [x] Code complies with PEP-8 and is consistent with the framework.
• [x] Code is properly documented.
• [x] PR description covers ALL the changes performed.
• [x] Files changed (update, add, delete) are in the PR's scope (no extra files are included).

Tests

• [x] New functionality is tested.
• [x] Tests are created for the new functionality or existing tests are updated accordingly.
• [x] ALL tests pass with no errors.
• [x] CI/CD pipelines run with no errors.
• [x] Test Coverage is maintained (coverage may drop by no more than 0.2%).

Expected behaviour

For the same code, the final results of Hoeffding Tree should be the same regardless of the version of python.

Actual behaviour

Hoeffding Tree returns a different probability for the same code, and this probability is the same in python 3.4 & 3.5 but different in 3.6. This behaviour doesn't depend on leaf_prediction.

Also, the classifier has the same model (tree) in the end for each version of python.

I did a little digging, and I found that the origin of the problem is in the final values of _observed_class_distribution of active nodes.

Here is a quick example:

    stream = RandomTreeGenerator(tree_random_state=23, sample_random_state=12, n_classes=2, n_cat_features=2,
                                     n_categories_per_cat_feature=4, n_num_features=1, max_tree_depth=30,
                                     min_leaf_depth=10,
                                     fraction_leaves_per_level=0.45)
    stream.prepare_for_use()

    learner = HoeffdingTree(nominal_attributes=[i for i in range(1, 9)], leaf_prediction='nba')

    cnt = 0
    max_samples = 15000
    predictions = array('i')
    proba_predictions = []
    wait_samples = 100

    while cnt < max_samples:
        X, y = stream.next_sample()
        # Test every n samples
        if (cnt % wait_samples == 0) and (cnt != 0):
            predictions.append(learner.predict(X)[0])
            proba_predictions.append(learner.predict_proba(X)[0])
        learner.partial_fit(X, y)
        cnt += 1

    l_nodes = learner._find_learning_nodes()
    for node in l_nodes:
        print(node.node._observed_class_distribution.values())
        

# python3.6

dict_values([592.0, 535.0])
dict_values([227.0, 304.0])
dict_values([830.0, 542.0])
dict_values([1594.0, 1623.0])
dict_values([324.79304610816257, 582.6026022538947])
dict_values([822.2069538918374, 1131.3973977461053])

# python3.4 & 3.5
dict_values([587.0, 640.0])
dict_values([327.0, 254.0])
dict_values([830.0, 542.0])
dict_values([1414.0, 1529.0])
dict_values([582.6026022538947, 324.79304610816257])
dict_values([1131.3973977461053, 822.2069538918374])

[Note: For comments and usage-related questions please consider the users group.

The running time should be included as an evaluation metric. It is already being logged for the overall process, but it would be nice to get a plot over time. It could also be interesting to split into training time / testing time.

Describe the bug Hi everyone! ,

I am working on drift detection using ADWIN. My concern is very specific to detected_change() method while understanding the code I observe that inner-most for loop "k" is iterated over range(cursor.bucket_size_row - 1). Since bucket_size_row indicates the number of elements present in the bucket and iteration of the inner-most loop in detected_change() we are skipping the last non-zero elements while comparing the absolute value with threshold epsilon. I might be wrong but wanted to understand the reasoning and related concepts explained in the research paper. Please advise on the reasoning and logic of detected_change() method.

  By Example: Let's assume by adding elements below are buckets being formed - 3rd Cursor Total : [ 3. 1. 96. 92. 0. 0.] 2nd Cursor Total : [2. 4. 2. 3. 0. 0.] 1st Cursor Total : [8. 4. 5. 9. 4. 0.]

What is happening in the current scenario - 5th Element (4) of 1st cursor total bucket is skipped to the first element (2) of 2nd cursor total bucket 4th Element (3) of 1st cursor total bucket is skipped to the first element(3) of 3rd cursor total bucket

Thanks in advance!

Additional context

def detected_change(self): """ Detects concept change in a drifting data stream.

    The ADWIN algorithm is described in Bifet and Gavaldà's 'Learning from 
    Time-Changing Data with Adaptive Windowing'. The general idea is to keep 
    statistics from a window of variable size while detecting concept drift.
     
    This function is responsible for analysing different cutting points in 
    the sliding window, to verify if there is a significant change in concept.
    
    Returns
    -------
    bln_change : bool
        Whether change was detected or not
        
    Notes
    -----
    If change was detected, one should verify the new window size, by reading 
    the width property.
    
    """
    bln_change = False
    bln_exit = False
    bln_bucket_deleted = False
    self.mint_time += 1
    n0 = 0
    if (self.mint_time % self.mint_clock == 0) and (self.width > self.mint_min_window_longitude):
        bln_reduce_width = True
        while bln_reduce_width:
            bln_reduce_width = not bln_reduce_width
            bln_exit = False
            n0 = 0
            n1 = self._width
            u0 = 0
            u1 = self.total
            v0 = 0
            v1 = self._variance
            n2 = 0
            u2 = 0
            cursor = self.list_row_bucket.last
            i = self.last_bucket_row

            while (not bln_exit) and (cursor is not None):
                for k in range(cursor.bucket_size_row - 1):
                    n2 = self.bucket_size(i)
                    u2 = cursor.get_total(k)

                    if n0 > 0:
                        v0 += cursor.get_variance(k) + 1. * n0 * n2 * (u0/n0 - u2/n2) * (u0/n0 - u2/n2) / (n0 + n2)

                    if n1 > 0:
                        v1 -= cursor.get_variance(k) + 1. * n1 * n2 * (u1/n1 - u2/n2) * (u1/n1 - u2/n2) / (n1 + n2)

                    n0 += self.bucket_size(i)
                    n1 -= self.bucket_size(i)
                    u0 += cursor.get_total(k)
                    u1 -= cursor.get_total(k)

                    if (i == 0) and (k == cursor.bucket_size_row - 1):
                        bln_exit = True
                        break

                    abs_value = 1. * ((u0/n0) - (u1/n1))
                    if (n1 >= self.mint_min_window_length) and (n0 >= self.mint_min_window_length)\
                            and (self.__bln_cut_expression(n0, n1, u0, u1, v0, v1, abs_value, self.delta)):
                        bln_bucket_deleted = True
                        self.detect = self.mint_time
                        if self.detect == 0:
                            self.detect = self.mint_time
                        elif self.detect_twice == 0:
                            self.detect_twice = self.mint_time

                        bln_reduce_width = True
                        bln_change = True
                        if self.width > 0:
                            n0 -= self.delete_element()
                            bln_exit = True
                            break

                cursor = cursor.get_previous()
                i -= 1
    self.mdbl_width += self.width
    if bln_change:
        self._n_detections += 1
    self.in_concept_change = bln_change
    return bln_change

@jacobmontiel

Bumps numpy from 1.17.3 to 1.22.0.

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

There are two differences comparing the MOA and scikit-multiflow implementation detected and fixed:

1. When a concept drift is detected, the statistics should be reset. If not, the detector will continuously detect more drifts after the first detection. One example is the “drift_stream.npy” applied in the unit test. The stream contains one drift that starts at index 999 (the distribution of values changed). However, with the current implementation, four drifts are detected

With the current source code:

After resetting the statistics on the add_element method:

2. In the method detected_change, the code does not iterate over the complete bucket because of this line: for k in range(cursor.bucket_size_row-1):

I have changed the code line to solve it because the range function does not include the cursor value.bucket_size_row. New code line: for k in range(cursor.bucket_size_row):

Changes proposed in this pull request:

• Change method add_element() from ADWIN class to reset statistics if a concept drift was detected.
• Change the method detected_change from ADWIN class to iterate the complete bucket.

Checklist

Implementation

• [ ] Implementation is correct (it performs its intended function).
• [ ] Code is consistent with the framework.
• [ ] Code is properly documented.
• [ ] PR description covers ALL the changes performed.
• [ ] Files changed (update, add, delete) are in the PR's scope (no extra files are included).

Tests

• [X ] New functionality is tested.
• [X ] Tests are created for the new functionality or existing tests are updated accordingly.
• [X ] ALL tests pass with no errors.
• [ ] CI/CD pipelines run with no errors.
• [ ] Test Coverage is maintained (coverage may drop by no more than 0.2%).

Is your feature request related to a problem? Please describe. I have a training dataset of around 1.5m records. I was trying to get HoeffdingTreeRegressor to fit it, and it's been running for a while without any indication about its progress.

Describe the solution you'd like It'd be great to have verbose: int param in the constructor to report what's happening within the fitting process based on the level (in int) passed to it.

Describe alternatives you've considered Not really any alternative.

Additional context E.g. following regressors all accept verbose param.

ensemble.RandomForestRegressor(n_jobs=-1, random_state=rand_state, verbose=1)
ensemble.BaggingRegressor(n_jobs=-1, random_state=rand_state, verbose=1)
xgb.XGBRegressor(verbosity=1, booster='gbtree', n_jobs=-1, random_state=rand_state)
lgb.LGBMRegressor(num_leaves=2047, random_state=rand_state, force_col_wise=True, verbose=1)

I'm trying to install scikit-multiflow by following the installation guide on the webpage. I've installed the dependencies but I can't install it using the pip command "pip install -U scikit-multiflow". Could you please help me?

The following is the installation log:

ERROR: Command errored out with exit status 1: command: 'c:\users\pichau\pycharmprojects\tccfwork\venv\scripts\python.exe' -u -c 'import io, os, sys, setuptools , tokenize; sys.argv[0] = '"'"'C:\Users\Pichau\AppData\Local\Temp\pip-install-wxbtl444\scikit-multiflow_003d6 117f70b419bb0cf4433d5182b7c\setup.py'"'"'; file='"'"'C:\Users\Pichau\AppData\Local\Temp\pip-install-wxbtl 444\scikit-multiflow_003d6117f70b419bb0cf4433d5182b7c\setup.py'"'"';f = getattr(tokenize, '"'"'open'"'"', open)(__ file__) if os.path.exists(file) else io.StringIO('"'"'from setuptools import setup; setup()'"'"');code = f.read( ).replace('"'"'\r\n'"'"', '"'"'\n'"'"');f.close();exec(compile(code, file, '"'"'exec'"'"'))' bdist_wheel -d 'C:
Users\Pichau\AppData\Local\Temp\pip-wheel-2qm77rhr' cwd: C:\Users\Pichau\AppData\Local\Temp\pip-install-wxbtl444\scikit-multiflow_003d6117f70b419bb0cf4433d5182b7 c
Complete output (247 lines): running bdist_wheel running build running config_cc unifing config_cc, config, build_clib, build_ext, build commands --compiler options running config_fc unifing config_fc, config, build_clib, build_ext, build commands --fcompiler options running build_src build_src building extension "skmultiflow.lazy.libNearestNeighbor" sources building extension "skmultiflow.metrics.confusion_matrix" sources building extension "skmultiflow.metrics.classification_performance_evaluator" sources build_src: building npy-pkg config files running build_py creating build creating build\lib.win32-3.6 creating build\lib.win32-3.6\skmultiflow copying src\skmultiflow\setup.py -> build\lib.win32-3.6\skmultiflow copying src\skmultiflow_version.py -> build\lib.win32-3.6\skmultiflow copying src\skmultiflow_init.py -> build\lib.win32-3.6\skmultiflow creating build\lib.win32-3.6\skmultiflow\anomaly_detection copying src\skmultiflow\anomaly_detection\half_space_trees.py -> build\lib.win32-3.6\skmultiflow\anomaly_detection copying src\skmultiflow\anomaly_detection_init.py -> build\lib.win32-3.6\skmultiflow\anomaly_detection creating build\lib.win32-3.6\skmultiflow\bayes copying src\skmultiflow\bayes\naive_bayes.py -> build\lib.win32-3.6\skmultiflow\bayes copying src\skmultiflow\bayes\utils.py -> build\lib.win32-3.6\skmultiflow\bayes copying src\skmultiflow\bayes_init_.py -> build\lib.win32-3.6\skmultiflow\bayes creating build\lib.win32-3.6\skmultiflow\core copying src\skmultiflow\core\base.py -> build\lib.win32-3.6\skmultiflow\core copying src\skmultiflow\core\instance_header.py -> build\lib.win32-3.6\skmultiflow\core copying src\skmultiflow\core\pipeline.py -> build\lib.win32-3.6\skmultiflow\core copying src\skmultiflow\core_init_.py -> build\lib.win32-3.6\skmultiflow\core creating build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\agrawal_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\anomaly_sine_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\base_stream.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\concept_drift_stream.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\data_stream.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\file_stream.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\hyper_plane_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\led_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\led_generator_drift.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\mixed_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\multilabel_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\pseudo_random_processes.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\random_rbf_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\random_rbf_generator_drift.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\random_tree_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\regression_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\sea_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\sine_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\stagger_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\synth.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\temporal_data_stream.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\time_manager.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\unsupervised_stream.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\waveform_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data_init_.py -> build\lib.win32-3.6\skmultiflow\data creating build\lib.win32-3.6\skmultiflow\drift_detection copying src\skmultiflow\drift_detection\adwin.py -> build\lib.win32-3.6\skmultiflow\drift_detection copying src\skmultiflow\drift_detection\base_drift_detector.py -> build\lib.win32-3.6\skmultiflow\drift_detection copying src\skmultiflow\drift_detection\ddm.py -> build\lib.win32-3.6\skmultiflow\drift_detection copying src\skmultiflow\drift_detection\eddm.py -> build\lib.win32-3.6\skmultiflow\drift_detection copying src\skmultiflow\drift_detection\hddm_a.py -> build\lib.win32-3.6\skmultiflow\drift_detection copying src\skmultiflow\drift_detection\hddm_w.py -> build\lib.win32-3.6\skmultiflow\drift_detection copying src\skmultiflow\drift_detection\kswin.py -> build\lib.win32-3.6\skmultiflow\drift_detection copying src\skmultiflow\drift_detection\page_hinkley.py -> build\lib.win32-3.6\skmultiflow\drift_detection copying src\skmultiflow\drift_detection_init_.py -> build\lib.win32-3.6\skmultiflow\drift_detection creating build\lib.win32-3.6\skmultiflow\evaluation copying src\skmultiflow\evaluation\base_evaluator.py -> build\lib.win32-3.6\skmultiflow\evaluation copying src\skmultiflow\evaluation\evaluate_holdout.py -> build\lib.win32-3.6\skmultiflow\evaluation copying src\skmultiflow\evaluation\evaluate_prequential.py -> build\lib.win32-3.6\skmultiflow\evaluation copying src\skmultiflow\evaluation\evaluate_prequential_delayed.py -> build\lib.win32-3.6\skmultiflow\evaluation copying src\skmultiflow\evaluation\evaluate_stream_gen_speed.py -> build\lib.win32-3.6\skmultiflow\evaluation copying src\skmultiflow\evaluation\evaluation_data_buffer.py -> build\lib.win32-3.6\skmultiflow\evaluation copying src\skmultiflow\evaluation_init_.py -> build\lib.win32-3.6\skmultiflow\evaluation creating build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\accuracy_weighted_ensemble.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\adaptive_random_forests.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\adaptive_random_forest_regressor.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\additive_expert_ensemble.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\batch_incremental.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\classifier_chains.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\dynamic_weighted_majority.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\learn_nse.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\learn_pp.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\leverage_bagging.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\multi_output_learner.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\online_adac2.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\online_boosting.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\online_csb2.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\online_rus_boost.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\online_smote_bagging.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\online_under_over_bagging.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\oza_bagging.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\oza_bagging_adwin.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\regressor_chains.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\streaming_random_patches.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta_init_.py -> build\lib.win32-3.6\skmultiflow\meta creating build\lib.win32-3.6\skmultiflow\neural_networks copying src\skmultiflow\neural_networks\perceptron.py -> build\lib.win32-3.6\skmultiflow\neural_networks copying src\skmultiflow\neural_networks_init_.py -> build\lib.win32-3.6\skmultiflow\neural_networks creating build\lib.win32-3.6\skmultiflow\prototype copying src\skmultiflow\prototype\robust_soft_learning_vector_quantization.py -> build\lib.win32-3.6\skmultiflow\p rototype copying src\skmultiflow\prototype_init_.py -> build\lib.win32-3.6\skmultiflow\prototype creating build\lib.win32-3.6\skmultiflow\rules copying src\skmultiflow\rules\attribute_expand_suggestion.py -> build\lib.win32-3.6\skmultiflow\rules copying src\skmultiflow\rules\base_predicate.py -> build\lib.win32-3.6\skmultiflow\rules copying src\skmultiflow\rules\base_rule.py -> build\lib.win32-3.6\skmultiflow\rules copying src\skmultiflow\rules\foil_gain_rule_criterion.py -> build\lib.win32-3.6\skmultiflow\rules copying src\skmultiflow\rules\hellinger_distance_criterion.py -> build\lib.win32-3.6\skmultiflow\rules copying src\skmultiflow\rules\info_gain_rule_criterion.py -> build\lib.win32-3.6\skmultiflow\rules copying src\skmultiflow\rules\nominal_attribute_class_observer.py -> build\lib.win32-3.6\skmultiflow\rules copying src\skmultiflow\rules\numeric_attribute_class_observer.py -> build\lib.win32-3.6\skmultiflow\rules copying src\skmultiflow\rules\very_fast_decision_rules.py -> build\lib.win32-3.6\skmultiflow\rules copying src\skmultiflow\rules_init_.py -> build\lib.win32-3.6\skmultiflow\rules creating build\lib.win32-3.6\skmultiflow\transform copying src\skmultiflow\transform\base_transform.py -> build\lib.win32-3.6\skmultiflow\transform copying src\skmultiflow\transform\missing_values_cleaner.py -> build\lib.win32-3.6\skmultiflow\transform copying src\skmultiflow\transform\one_hot_to_categorical.py -> build\lib.win32-3.6\skmultiflow\transform copying src\skmultiflow\transform\windowed_minmax_scaler.py -> build\lib.win32-3.6\skmultiflow\transform copying src\skmultiflow\transform\windowed_standard_scaler.py -> build\lib.win32-3.6\skmultiflow\transform copying src\skmultiflow\transform_init_.py -> build\lib.win32-3.6\skmultiflow\transform creating build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\arf_hoeffding_tree.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\arf_hoeffding_tree_regressor.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\attribute_split_suggestion.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\extremely_fast_decision_tree.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\gaussian_estimator.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\hoeffding_adaptive_tree.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\hoeffding_adaptive_tree_regressor.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\hoeffding_tree.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\hoeffding_tree_regressor.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\isoup_tree.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\label_combination_hoeffding_tree.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\setup.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\stacked_single_target_hoeffding_tree_regressor.py -> build\lib.win32-3.6\skmultiflow \trees copying src\skmultiflow\trees_init_.py -> build\lib.win32-3.6\skmultiflow\trees creating build\lib.win32-3.6\skmultiflow\trees\attribute_observer copying src\skmultiflow\trees\attribute_observer\attribute_class_observer.py -> build\lib.win32-3.6\skmultiflow\tr ees\attribute_observer copying src\skmultiflow\trees\attribute_observer\attribute_class_observer_null.py -> build\lib.win32-3.6\skmultifl ow\trees\attribute_observer copying src\skmultiflow\trees\attribute_observer\nominal_attribute_class_observer.py -> build\lib.win32-3.6\skmult iflow\trees\attribute_observer copying src\skmultiflow\trees\attribute_observer\nominal_attribute_regression_observer.py -> build\lib.win32-3.6\s kmultiflow\trees\attribute_observer copying src\skmultiflow\trees\attribute_observer\numeric_attribute_class_observer_binary_tree.py -> build\lib.win3 2-3.6\skmultiflow\trees\attribute_observer copying src\skmultiflow\trees\attribute_observer\numeric_attribute_class_observer_gaussian.py -> build\lib.win32-3 .6\skmultiflow\trees\attribute_observer copying src\skmultiflow\trees\attribute_observer\numeric_attribute_regression_observer.py -> build\lib.win32-3.6\s kmultiflow\trees\attribute_observer copying src\skmultiflow\trees\attribute_observer_init_.py -> build\lib.win32-3.6\skmultiflow\trees\attribute_ob server creating build\lib.win32-3.6\skmultiflow\trees\attribute_test copying src\skmultiflow\trees\attribute_test\instance_conditional_test.py -> build\lib.win32-3.6\skmultiflow\trees \attribute_test copying src\skmultiflow\trees\attribute_test\nominal_attribute_binary_test.py -> build\lib.win32-3.6\skmultiflow\t rees\attribute_test copying src\skmultiflow\trees\attribute_test\nominal_attribute_multiway_test.py -> build\lib.win32-3.6\skmultiflow \trees\attribute_test copying src\skmultiflow\trees\attribute_test\numeric_attribute_binary_test.py -> build\lib.win32-3.6\skmultiflow\t rees\attribute_test copying src\skmultiflow\trees\attribute_test_init_.py -> build\lib.win32-3.6\skmultiflow\trees\attribute_test creating build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\active_learning_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\active_learning_node_adaptive_multi_target.py -> build\lib.win32-3.6\skmultifl ow\trees\nodes copying src\skmultiflow\trees\nodes\active_learning_node_for_regression.py -> build\lib.win32-3.6\skmultiflow\tree s\nodes copying src\skmultiflow\trees\nodes\active_learning_node_for_regression_multi_target.py -> build\lib.win32-3.6\skm ultiflow\trees\nodes copying src\skmultiflow\trees\nodes\active_learning_node_perceptron.py -> build\lib.win32-3.6\skmultiflow\trees\no des copying src\skmultiflow\trees\nodes\active_learning_node_perceptron_multi_target.py -> build\lib.win32-3.6\skmulti flow\trees\nodes copying src\skmultiflow\trees\nodes\ada_learning_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\ada_learning_node_for_regression.py -> build\lib.win32-3.6\skmultiflow\trees\n odes copying src\skmultiflow\trees\nodes\ada_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\ada_split_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\ada_split_node_for_regression.py -> build\lib.win32-3.6\skmultiflow\trees\node s copying src\skmultiflow\trees\nodes\anytime_active_learning_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\anytime_inactive_learning_node.py -> build\lib.win32-3.6\skmultiflow\trees\nod es copying src\skmultiflow\trees\nodes\anytime_learning_node_nb.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\anytime_learning_node_nb_adaptive.py -> build\lib.win32-3.6\skmultiflow\trees
nodes copying src\skmultiflow\trees\nodes\anytime_split_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\found_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\inactive_learning_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\inactive_learning_node_adaptive_multi_target.py -> build\lib.win32-3.6\skmulti flow\trees\nodes copying src\skmultiflow\trees\nodes\inactive_learning_node_for_regression.py -> build\lib.win32-3.6\skmultiflow\tr ees\nodes copying src\skmultiflow\trees\nodes\inactive_learning_node_perceptron.py -> build\lib.win32-3.6\skmultiflow\trees
nodes copying src\skmultiflow\trees\nodes\inactive_learning_node_perceptron_multi_target.py -> build\lib.win32-3.6\skmul tiflow\trees\nodes copying src\skmultiflow\trees\nodes\lc_active_learning_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\lc_inactive_learning_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\lc_learning_node_nb.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\lc_learning_node_nba.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\learning_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\learning_node_nb.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\learning_node_nb_adaptive.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\random_learning_node_classification.py -> build\lib.win32-3.6\skmultiflow\tree s\nodes copying src\skmultiflow\trees\nodes\random_learning_node_for_regression.py -> build\lib.win32-3.6\skmultiflow\tree s\nodes copying src\skmultiflow\trees\nodes\random_learning_node_nb.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\random_learning_node_nb_adaptive.py -> build\lib.win32-3.6\skmultiflow\trees\n odes copying src\skmultiflow\trees\nodes\random_learning_node_perceptron.py -> build\lib.win32-3.6\skmultiflow\trees\no des copying src\skmultiflow\trees\nodes\split_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\sst_active_learning_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\sst_active_learning_node_adaptive.py -> build\lib.win32-3.6\skmultiflow\trees
nodes copying src\skmultiflow\trees\nodes\sst_inactive_learning_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\sst_inactive_learning_node_adaptive.py -> build\lib.win32-3.6\skmultiflow\tree s\nodes copying src\skmultiflow\trees\nodes_init_.py -> build\lib.win32-3.6\skmultiflow\trees\nodes creating build\lib.win32-3.6\skmultiflow\trees\split_criterion copying src\skmultiflow\trees\split_criterion\base_split_criterion.py -> build\lib.win32-3.6\skmultiflow\trees\spl it_criterion copying src\skmultiflow\trees\split_criterion\gini_split_criterion.py -> build\lib.win32-3.6\skmultiflow\trees\spl it_criterion copying src\skmultiflow\trees\split_criterion\hellinger_distance_criterion.py -> build\lib.win32-3.6\skmultiflow\t rees\split_criterion copying src\skmultiflow\trees\split_criterion\info_gain_split_criterion.py -> build\lib.win32-3.6\skmultiflow\tree s\split_criterion copying src\skmultiflow\trees\split_criterion\intra_cluster_variance_reduction_split_criterion.py -> build\lib.win 32-3.6\skmultiflow\trees\split_criterion copying src\skmultiflow\trees\split_criterion\variance_reduction_split_criterion.py -> build\lib.win32-3.6\skmulti flow\trees\split_criterion copying src\skmultiflow\trees\split_criterion_init_.py -> build\lib.win32-3.6\skmultiflow\trees\split_criterion creating build\lib.win32-3.6\skmultiflow\utils copying src\skmultiflow\utils\constants.py -> build\lib.win32-3.6\skmultiflow\utils copying src\skmultiflow\utils\data_structures.py -> build\lib.win32-3.6\skmultiflow\utils copying src\skmultiflow\utils\file_scripts.py -> build\lib.win32-3.6\skmultiflow\utils copying src\skmultiflow\utils\statistics.py -> build\lib.win32-3.6\skmultiflow\utils copying src\skmultiflow\utils\utils.py -> build\lib.win32-3.6\skmultiflow\utils copying src\skmultiflow\utils\validation.py -> build\lib.win32-3.6\skmultiflow\utils copying src\skmultiflow\utils_pprint.py -> build\lib.win32-3.6\skmultiflow\utils copying src\skmultiflow\utils_show_versions.py -> build\lib.win32-3.6\skmultiflow\utils copying src\skmultiflow\utils_init_.py -> build\lib.win32-3.6\skmultiflow\utils creating build\lib.win32-3.6\skmultiflow\visualization copying src\skmultiflow\visualization\base_listener.py -> build\lib.win32-3.6\skmultiflow\visualization copying src\skmultiflow\visualization\evaluation_visualizer.py -> build\lib.win32-3.6\skmultiflow\visualization copying src\skmultiflow\visualization_init_.py -> build\lib.win32-3.6\skmultiflow\visualization creating build\lib.win32-3.6\skmultiflow\lazy copying src\skmultiflow\lazy\base_neighbors.py -> build\lib.win32-3.6\skmultiflow\lazy copying src\skmultiflow\lazy\knn_adwin.py -> build\lib.win32-3.6\skmultiflow\lazy copying src\skmultiflow\lazy\knn_classifier.py -> build\lib.win32-3.6\skmultiflow\lazy copying src\skmultiflow\lazy\knn_regressor.py -> build\lib.win32-3.6\skmultiflow\lazy copying src\skmultiflow\lazy\sam_knn.py -> build\lib.win32-3.6\skmultiflow\lazy copying src\skmultiflow\lazy\setup.py -> build\lib.win32-3.6\skmultiflow\lazy copying src\skmultiflow\lazy_init_.py -> build\lib.win32-3.6\skmultiflow\lazy creating build\lib.win32-3.6\skmultiflow\metrics copying src\skmultiflow\metrics\measure_collection.py -> build\lib.win32-3.6\skmultiflow\metrics copying src\skmultiflow\metrics\setup.py -> build\lib.win32-3.6\skmultiflow\metrics copying src\skmultiflow\metrics_init_.py -> build\lib.win32-3.6\skmultiflow\metrics running build_ext No module named 'numpy.distutils._msvccompiler' in numpy.distutils; trying from distutils customize MSVCCompiler customize MSVCCompiler using build_ext No module named 'numpy.distutils._msvccompiler' in numpy.distutils; trying from distutils customize MSVCCompiler Missing compiler_cxx fix for MSVCCompiler customize MSVCCompiler using build_ext building 'skmultiflow.lazy.libNearestNeighbor' extension compiling C sources creating build\temp.win32-3.6\Release\src creating build\temp.win32-3.6\Release\src\skmultiflow creating build\temp.win32-3.6\Release\src\skmultiflow\lazy creating build\temp.win32-3.6\Release\src\skmultiflow\lazy\src creating build\temp.win32-3.6\Release\src\skmultiflow\lazy\src\libNearestNeighbors C:\Program Files (x86)\Microsoft Visual Studio\2019\Community\VC\Tools\MSVC\14.27.29110\bin\HostX86\x86\cl.exe /c /nologo /Ox /W3 /GL /DNDEBUG /MD -Ic:\users\pichau\pycharmprojects\tccfwork\venv\lib\site-packages\numpy\core\includ e -Ic:\users\pichau\pycharmprojects\tccfwork\venv\lib\site-packages\numpy\core\include -Ic:\users\pichau\pycharmproj ects\tccfwork\venv\include -IC:\Users\Pichau\AppData\Local\Programs\Python\Python36-32\include -IC:\Users\Pichau\App Data\Local\Programs\Python\Python36-32\include -IC:\Program Files (x86)\Microsoft Visual Studio\2019\Community\VC\To ols\MSVC\14.27.29110\include /EHsc /Tpsrc\skmultiflow\lazy\src\libNearestNeighbors\nearestNeighbor.cpp /Fobuild\temp .win32-3.6\Release\src\skmultiflow\lazy\src\libNearestNeighbors\nearestNeighbor.obj /Zm1000 nearestNeighbor.cpp C:\Users\Pichau\AppData\Local\Programs\Python\Python36-32\include\pyconfig.h(59): fatal error C1083: NÆo ‚ poss¡ve l abrir arquivo incluir: 'io.h': No such file or directory error: Command "C:\Program Files (x86)\Microsoft Visual Studio\2019\Community\VC\Tools\MSVC\14.27.29110\bin\HostX8 6\x86\cl.exe /c /nologo /Ox /W3 /GL /DNDEBUG /MD -Ic:\users\pichau\pycharmprojects\tccfwork\venv\lib\site-packages\n umpy\core\include -Ic:\users\pichau\pycharmprojects\tccfwork\venv\lib\site-packages\numpy\core\include -Ic:\users\pi chau\pycharmprojects\tccfwork\venv\include -IC:\Users\Pichau\AppData\Local\Programs\Python\Python36-32\include -IC:
Users\Pichau\AppData\Local\Programs\Python\Python36-32\include -IC:\Program Files (x86)\Microsoft Visual Studio\2019 \Community\VC\Tools\MSVC\14.27.29110\include /EHsc /Tpsrc\skmultiflow\lazy\src\libNearestNeighbors\nearestNeighbor.c pp /Fobuild\temp.win32-3.6\Release\src\skmultiflow\lazy\src\libNearestNeighbors\nearestNeighbor.obj /Zm1000" failed with exit status 2

ERROR: Failed building wheel for scikit-multiflow Running setup.py clean for scikit-multiflow Failed to build scikit-multiflow Installing collected packages: scikit-multiflow Running setup.py install for scikit-multiflow ... error ERROR: Command errored out with exit status 1: command: 'c:\users\pichau\pycharmprojects\tccfwork\venv\scripts\python.exe' -u -c 'import io, os, sys, setuptoo ls, tokenize; sys.argv[0] = '"'"'C:\Users\Pichau\AppData\Local\Temp\pip-install-wxbtl444\scikit-multiflow_003 d6117f70b419bb0cf4433d5182b7c\setup.py'"'"'; file='"'"'C:\Users\Pichau\AppData\Local\Temp\pip-install-wxb tl444\scikit-multiflow_003d6117f70b419bb0cf4433d5182b7c\setup.py'"'"';f = getattr(tokenize, '"'"'open'"'"', open)( file) if os.path.exists(file) else io.StringIO('"'"'from setuptools import setup; setup()'"'"');code = f.rea d().replace('"'"'\r\n'"'"', '"'"'\n'"'"');f.close();exec(compile(code, file, '"'"'exec'"'"'))' install --record 'C:\Users\Pichau\AppData\Local\Temp\pip-record-tsuf4uvp\install-record.txt' --single-version-externally-managed --co mpile --install-headers 'c:\users\pichau\pycharmprojects\tccfwork\venv\include\site\python3.6\scikit-multiflow' cwd: C:\Users\Pichau\AppData\Local\Temp\pip-install-wxbtl444\scikit-multiflow_003d6117f70b419bb0cf4433d5182 b7c
Complete output (247 lines): running install running build running config_cc unifing config_cc, config, build_clib, build_ext, build commands --compiler options running config_fc unifing config_fc, config, build_clib, build_ext, build commands --fcompiler options running build_src build_src building extension "skmultiflow.lazy.libNearestNeighbor" sources building extension "skmultiflow.metrics.confusion_matrix" sources building extension "skmultiflow.metrics.classification_performance_evaluator" sources build_src: building npy-pkg config files running build_py creating build creating build\lib.win32-3.6 creating build\lib.win32-3.6\skmultiflow copying src\skmultiflow\setup.py -> build\lib.win32-3.6\skmultiflow copying src\skmultiflow_version.py -> build\lib.win32-3.6\skmultiflow copying src\skmultiflow_init.py -> build\lib.win32-3.6\skmultiflow creating build\lib.win32-3.6\skmultiflow\anomaly_detection copying src\skmultiflow\anomaly_detection\half_space_trees.py -> build\lib.win32-3.6\skmultiflow\anomaly_detecti on copying src\skmultiflow\anomaly_detection_init.py -> build\lib.win32-3.6\skmultiflow\anomaly_detection creating build\lib.win32-3.6\skmultiflow\bayes copying src\skmultiflow\bayes\naive_bayes.py -> build\lib.win32-3.6\skmultiflow\bayes copying src\skmultiflow\bayes\utils.py -> build\lib.win32-3.6\skmultiflow\bayes copying src\skmultiflow\bayes_init_.py -> build\lib.win32-3.6\skmultiflow\bayes creating build\lib.win32-3.6\skmultiflow\core copying src\skmultiflow\core\base.py -> build\lib.win32-3.6\skmultiflow\core copying src\skmultiflow\core\instance_header.py -> build\lib.win32-3.6\skmultiflow\core copying src\skmultiflow\core\pipeline.py -> build\lib.win32-3.6\skmultiflow\core copying src\skmultiflow\core_init_.py -> build\lib.win32-3.6\skmultiflow\core creating build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\agrawal_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\anomaly_sine_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\base_stream.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\concept_drift_stream.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\data_stream.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\file_stream.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\hyper_plane_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\led_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\led_generator_drift.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\mixed_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\multilabel_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\pseudo_random_processes.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\random_rbf_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\random_rbf_generator_drift.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\random_tree_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\regression_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\sea_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\sine_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\stagger_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\synth.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\temporal_data_stream.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\time_manager.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\unsupervised_stream.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data\waveform_generator.py -> build\lib.win32-3.6\skmultiflow\data copying src\skmultiflow\data_init_.py -> build\lib.win32-3.6\skmultiflow\data creating build\lib.win32-3.6\skmultiflow\drift_detection copying src\skmultiflow\drift_detection\adwin.py -> build\lib.win32-3.6\skmultiflow\drift_detection copying src\skmultiflow\drift_detection\base_drift_detector.py -> build\lib.win32-3.6\skmultiflow\drift_detectio n copying src\skmultiflow\drift_detection\ddm.py -> build\lib.win32-3.6\skmultiflow\drift_detection copying src\skmultiflow\drift_detection\eddm.py -> build\lib.win32-3.6\skmultiflow\drift_detection copying src\skmultiflow\drift_detection\hddm_a.py -> build\lib.win32-3.6\skmultiflow\drift_detection copying src\skmultiflow\drift_detection\hddm_w.py -> build\lib.win32-3.6\skmultiflow\drift_detection copying src\skmultiflow\drift_detection\kswin.py -> build\lib.win32-3.6\skmultiflow\drift_detection copying src\skmultiflow\drift_detection\page_hinkley.py -> build\lib.win32-3.6\skmultiflow\drift_detection copying src\skmultiflow\drift_detection_init_.py -> build\lib.win32-3.6\skmultiflow\drift_detection creating build\lib.win32-3.6\skmultiflow\evaluation copying src\skmultiflow\evaluation\base_evaluator.py -> build\lib.win32-3.6\skmultiflow\evaluation copying src\skmultiflow\evaluation\evaluate_holdout.py -> build\lib.win32-3.6\skmultiflow\evaluation copying src\skmultiflow\evaluation\evaluate_prequential.py -> build\lib.win32-3.6\skmultiflow\evaluation copying src\skmultiflow\evaluation\evaluate_prequential_delayed.py -> build\lib.win32-3.6\skmultiflow\evaluation copying src\skmultiflow\evaluation\evaluate_stream_gen_speed.py -> build\lib.win32-3.6\skmultiflow\evaluation copying src\skmultiflow\evaluation\evaluation_data_buffer.py -> build\lib.win32-3.6\skmultiflow\evaluation copying src\skmultiflow\evaluation_init_.py -> build\lib.win32-3.6\skmultiflow\evaluation creating build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\accuracy_weighted_ensemble.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\adaptive_random_forests.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\adaptive_random_forest_regressor.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\additive_expert_ensemble.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\batch_incremental.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\classifier_chains.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\dynamic_weighted_majority.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\learn_nse.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\learn_pp.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\leverage_bagging.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\multi_output_learner.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\online_adac2.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\online_boosting.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\online_csb2.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\online_rus_boost.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\online_smote_bagging.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\online_under_over_bagging.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\oza_bagging.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\oza_bagging_adwin.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\regressor_chains.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta\streaming_random_patches.py -> build\lib.win32-3.6\skmultiflow\meta copying src\skmultiflow\meta_init_.py -> build\lib.win32-3.6\skmultiflow\meta creating build\lib.win32-3.6\skmultiflow\neural_networks copying src\skmultiflow\neural_networks\perceptron.py -> build\lib.win32-3.6\skmultiflow\neural_networks copying src\skmultiflow\neural_networks_init_.py -> build\lib.win32-3.6\skmultiflow\neural_networks creating build\lib.win32-3.6\skmultiflow\prototype copying src\skmultiflow\prototype\robust_soft_learning_vector_quantization.py -> build\lib.win32-3.6\skmultiflow \prototype copying src\skmultiflow\prototype_init_.py -> build\lib.win32-3.6\skmultiflow\prototype creating build\lib.win32-3.6\skmultiflow\rules copying src\skmultiflow\rules\attribute_expand_suggestion.py -> build\lib.win32-3.6\skmultiflow\rules copying src\skmultiflow\rules\base_predicate.py -> build\lib.win32-3.6\skmultiflow\rules copying src\skmultiflow\rules\base_rule.py -> build\lib.win32-3.6\skmultiflow\rules copying src\skmultiflow\rules\foil_gain_rule_criterion.py -> build\lib.win32-3.6\skmultiflow\rules copying src\skmultiflow\rules\hellinger_distance_criterion.py -> build\lib.win32-3.6\skmultiflow\rules copying src\skmultiflow\rules\info_gain_rule_criterion.py -> build\lib.win32-3.6\skmultiflow\rules copying src\skmultiflow\rules\nominal_attribute_class_observer.py -> build\lib.win32-3.6\skmultiflow\rules copying src\skmultiflow\rules\numeric_attribute_class_observer.py -> build\lib.win32-3.6\skmultiflow\rules copying src\skmultiflow\rules\very_fast_decision_rules.py -> build\lib.win32-3.6\skmultiflow\rules copying src\skmultiflow\rules_init_.py -> build\lib.win32-3.6\skmultiflow\rules creating build\lib.win32-3.6\skmultiflow\transform copying src\skmultiflow\transform\base_transform.py -> build\lib.win32-3.6\skmultiflow\transform copying src\skmultiflow\transform\missing_values_cleaner.py -> build\lib.win32-3.6\skmultiflow\transform copying src\skmultiflow\transform\one_hot_to_categorical.py -> build\lib.win32-3.6\skmultiflow\transform copying src\skmultiflow\transform\windowed_minmax_scaler.py -> build\lib.win32-3.6\skmultiflow\transform copying src\skmultiflow\transform\windowed_standard_scaler.py -> build\lib.win32-3.6\skmultiflow\transform copying src\skmultiflow\transform_init_.py -> build\lib.win32-3.6\skmultiflow\transform creating build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\arf_hoeffding_tree.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\arf_hoeffding_tree_regressor.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\attribute_split_suggestion.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\extremely_fast_decision_tree.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\gaussian_estimator.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\hoeffding_adaptive_tree.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\hoeffding_adaptive_tree_regressor.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\hoeffding_tree.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\hoeffding_tree_regressor.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\isoup_tree.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\label_combination_hoeffding_tree.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\setup.py -> build\lib.win32-3.6\skmultiflow\trees copying src\skmultiflow\trees\stacked_single_target_hoeffding_tree_regressor.py -> build\lib.win32-3.6\skmultifl ow\trees copying src\skmultiflow\trees_init_.py -> build\lib.win32-3.6\skmultiflow\trees creating build\lib.win32-3.6\skmultiflow\trees\attribute_observer copying src\skmultiflow\trees\attribute_observer\attribute_class_observer.py -> build\lib.win32-3.6\skmultiflow
trees\attribute_observer copying src\skmultiflow\trees\attribute_observer\attribute_class_observer_null.py -> build\lib.win32-3.6\skmulti flow\trees\attribute_observer copying src\skmultiflow\trees\attribute_observer\nominal_attribute_class_observer.py -> build\lib.win32-3.6\skmu ltiflow\trees\attribute_observer copying src\skmultiflow\trees\attribute_observer\nominal_attribute_regression_observer.py -> build\lib.win32-3.6 \skmultiflow\trees\attribute_observer copying src\skmultiflow\trees\attribute_observer\numeric_attribute_class_observer_binary_tree.py -> build\lib.wi n32-3.6\skmultiflow\trees\attribute_observer copying src\skmultiflow\trees\attribute_observer\numeric_attribute_class_observer_gaussian.py -> build\lib.win32 -3.6\skmultiflow\trees\attribute_observer copying src\skmultiflow\trees\attribute_observer\numeric_attribute_regression_observer.py -> build\lib.win32-3.6 \skmultiflow\trees\attribute_observer copying src\skmultiflow\trees\attribute_observer_init_.py -> build\lib.win32-3.6\skmultiflow\trees\attribute_ observer creating build\lib.win32-3.6\skmultiflow\trees\attribute_test copying src\skmultiflow\trees\attribute_test\instance_conditional_test.py -> build\lib.win32-3.6\skmultiflow\tre es\attribute_test copying src\skmultiflow\trees\attribute_test\nominal_attribute_binary_test.py -> build\lib.win32-3.6\skmultiflow \trees\attribute_test copying src\skmultiflow\trees\attribute_test\nominal_attribute_multiway_test.py -> build\lib.win32-3.6\skmultifl ow\trees\attribute_test copying src\skmultiflow\trees\attribute_test\numeric_attribute_binary_test.py -> build\lib.win32-3.6\skmultiflow \trees\attribute_test copying src\skmultiflow\trees\attribute_test_init_.py -> build\lib.win32-3.6\skmultiflow\trees\attribute_test creating build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\active_learning_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\active_learning_node_adaptive_multi_target.py -> build\lib.win32-3.6\skmulti flow\trees\nodes copying src\skmultiflow\trees\nodes\active_learning_node_for_regression.py -> build\lib.win32-3.6\skmultiflow\tr ees\nodes copying src\skmultiflow\trees\nodes\active_learning_node_for_regression_multi_target.py -> build\lib.win32-3.6\s kmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\active_learning_node_perceptron.py -> build\lib.win32-3.6\skmultiflow\trees
nodes copying src\skmultiflow\trees\nodes\active_learning_node_perceptron_multi_target.py -> build\lib.win32-3.6\skmul tiflow\trees\nodes copying src\skmultiflow\trees\nodes\ada_learning_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\ada_learning_node_for_regression.py -> build\lib.win32-3.6\skmultiflow\trees \nodes copying src\skmultiflow\trees\nodes\ada_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\ada_split_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\ada_split_node_for_regression.py -> build\lib.win32-3.6\skmultiflow\trees\no des copying src\skmultiflow\trees\nodes\anytime_active_learning_node.py -> build\lib.win32-3.6\skmultiflow\trees\nod es copying src\skmultiflow\trees\nodes\anytime_inactive_learning_node.py -> build\lib.win32-3.6\skmultiflow\trees\n odes copying src\skmultiflow\trees\nodes\anytime_learning_node_nb.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\anytime_learning_node_nb_adaptive.py -> build\lib.win32-3.6\skmultiflow\tree s\nodes copying src\skmultiflow\trees\nodes\anytime_split_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\found_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\inactive_learning_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\inactive_learning_node_adaptive_multi_target.py -> build\lib.win32-3.6\skmul tiflow\trees\nodes copying src\skmultiflow\trees\nodes\inactive_learning_node_for_regression.py -> build\lib.win32-3.6\skmultiflow
trees\nodes copying src\skmultiflow\trees\nodes\inactive_learning_node_perceptron.py -> build\lib.win32-3.6\skmultiflow\tree s\nodes copying src\skmultiflow\trees\nodes\inactive_learning_node_perceptron_multi_target.py -> build\lib.win32-3.6\skm ultiflow\trees\nodes copying src\skmultiflow\trees\nodes\lc_active_learning_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\lc_inactive_learning_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\lc_learning_node_nb.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\lc_learning_node_nba.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\learning_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\learning_node_nb.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\learning_node_nb_adaptive.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\random_learning_node_classification.py -> build\lib.win32-3.6\skmultiflow\tr ees\nodes copying src\skmultiflow\trees\nodes\random_learning_node_for_regression.py -> build\lib.win32-3.6\skmultiflow\tr ees\nodes copying src\skmultiflow\trees\nodes\random_learning_node_nb.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\random_learning_node_nb_adaptive.py -> build\lib.win32-3.6\skmultiflow\trees \nodes copying src\skmultiflow\trees\nodes\random_learning_node_perceptron.py -> build\lib.win32-3.6\skmultiflow\trees
nodes copying src\skmultiflow\trees\nodes\split_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\sst_active_learning_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\sst_active_learning_node_adaptive.py -> build\lib.win32-3.6\skmultiflow\tree s\nodes copying src\skmultiflow\trees\nodes\sst_inactive_learning_node.py -> build\lib.win32-3.6\skmultiflow\trees\nodes copying src\skmultiflow\trees\nodes\sst_inactive_learning_node_adaptive.py -> build\lib.win32-3.6\skmultiflow\tr ees\nodes copying src\skmultiflow\trees\nodes_init_.py -> build\lib.win32-3.6\skmultiflow\trees\nodes creating build\lib.win32-3.6\skmultiflow\trees\split_criterion copying src\skmultiflow\trees\split_criterion\base_split_criterion.py -> build\lib.win32-3.6\skmultiflow\trees\s plit_criterion copying src\skmultiflow\trees\split_criterion\gini_split_criterion.py -> build\lib.win32-3.6\skmultiflow\trees\s plit_criterion copying src\skmultiflow\trees\split_criterion\hellinger_distance_criterion.py -> build\lib.win32-3.6\skmultiflow \trees\split_criterion copying src\skmultiflow\trees\split_criterion\info_gain_split_criterion.py -> build\lib.win32-3.6\skmultiflow\tr ees\split_criterion copying src\skmultiflow\trees\split_criterion\intra_cluster_variance_reduction_split_criterion.py -> build\lib.w in32-3.6\skmultiflow\trees\split_criterion copying src\skmultiflow\trees\split_criterion\variance_reduction_split_criterion.py -> build\lib.win32-3.6\skmul tiflow\trees\split_criterion copying src\skmultiflow\trees\split_criterion_init_.py -> build\lib.win32-3.6\skmultiflow\trees\split_criteri on creating build\lib.win32-3.6\skmultiflow\utils copying src\skmultiflow\utils\constants.py -> build\lib.win32-3.6\skmultiflow\utils copying src\skmultiflow\utils\data_structures.py -> build\lib.win32-3.6\skmultiflow\utils copying src\skmultiflow\utils\file_scripts.py -> build\lib.win32-3.6\skmultiflow\utils copying src\skmultiflow\utils\statistics.py -> build\lib.win32-3.6\skmultiflow\utils copying src\skmultiflow\utils\utils.py -> build\lib.win32-3.6\skmultiflow\utils copying src\skmultiflow\utils\validation.py -> build\lib.win32-3.6\skmultiflow\utils copying src\skmultiflow\utils_pprint.py -> build\lib.win32-3.6\skmultiflow\utils copying src\skmultiflow\utils_show_versions.py -> build\lib.win32-3.6\skmultiflow\utils copying src\skmultiflow\utils_init_.py -> build\lib.win32-3.6\skmultiflow\utils creating build\lib.win32-3.6\skmultiflow\visualization copying src\skmultiflow\visualization\base_listener.py -> build\lib.win32-3.6\skmultiflow\visualization copying src\skmultiflow\visualization\evaluation_visualizer.py -> build\lib.win32-3.6\skmultiflow\visualization copying src\skmultiflow\visualization_init_.py -> build\lib.win32-3.6\skmultiflow\visualization creating build\lib.win32-3.6\skmultiflow\lazy copying src\skmultiflow\lazy\base_neighbors.py -> build\lib.win32-3.6\skmultiflow\lazy copying src\skmultiflow\lazy\knn_adwin.py -> build\lib.win32-3.6\skmultiflow\lazy copying src\skmultiflow\lazy\knn_classifier.py -> build\lib.win32-3.6\skmultiflow\lazy copying src\skmultiflow\lazy\knn_regressor.py -> build\lib.win32-3.6\skmultiflow\lazy copying src\skmultiflow\lazy\sam_knn.py -> build\lib.win32-3.6\skmultiflow\lazy copying src\skmultiflow\lazy\setup.py -> build\lib.win32-3.6\skmultiflow\lazy copying src\skmultiflow\lazy_init_.py -> build\lib.win32-3.6\skmultiflow\lazy creating build\lib.win32-3.6\skmultiflow\metrics copying src\skmultiflow\metrics\measure_collection.py -> build\lib.win32-3.6\skmultiflow\metrics copying src\skmultiflow\metrics\setup.py -> build\lib.win32-3.6\skmultiflow\metrics copying src\skmultiflow\metrics_init_.py -> build\lib.win32-3.6\skmultiflow\metrics running build_ext No module named 'numpy.distutils._msvccompiler' in numpy.distutils; trying from distutils customize MSVCCompiler customize MSVCCompiler using build_ext No module named 'numpy.distutils._msvccompiler' in numpy.distutils; trying from distutils customize MSVCCompiler Missing compiler_cxx fix for MSVCCompiler customize MSVCCompiler using build_ext building 'skmultiflow.lazy.libNearestNeighbor' extension compiling C sources creating build\temp.win32-3.6\Release\src creating build\temp.win32-3.6\Release\src\skmultiflow creating build\temp.win32-3.6\Release\src\skmultiflow\lazy creating build\temp.win32-3.6\Release\src\skmultiflow\lazy\src creating build\temp.win32-3.6\Release\src\skmultiflow\lazy\src\libNearestNeighbors C:\Program Files (x86)\Microsoft Visual Studio\2019\Community\VC\Tools\MSVC\14.27.29110\bin\HostX86\x86\cl.exe / c /nologo /Ox /W3 /GL /DNDEBUG /MD -Ic:\users\pichau\pycharmprojects\tccfwork\venv\lib\site-packages\numpy\core\incl ude -Ic:\users\pichau\pycharmprojects\tccfwork\venv\lib\site-packages\numpy\core\include -Ic:\users\pichau\pycharmpr ojects\tccfwork\venv\include -IC:\Users\Pichau\AppData\Local\Programs\Python\Python36-32\include -IC:\Users\Pichau\A ppData\Local\Programs\Python\Python36-32\include -IC:\Program Files (x86)\Microsoft Visual Studio\2019\Community\VC
Tools\MSVC\14.27.29110\include /EHsc /Tpsrc\skmultiflow\lazy\src\libNearestNeighbors\nearestNeighbor.cpp /Fobuild\te mp.win32-3.6\Release\src\skmultiflow\lazy\src\libNearestNeighbors\nearestNeighbor.obj /Zm1000 nearestNeighbor.cpp C:\Users\Pichau\AppData\Local\Programs\Python\Python36-32\include\pyconfig.h(59): fatal error C1083: NÆo ‚ poss¡ vel abrir arquivo incluir: 'io.h': No such file or directory error: Command "C:\Program Files (x86)\Microsoft Visual Studio\2019\Community\VC\Tools\MSVC\14.27.29110\bin\Host X86\x86\cl.exe /c /nologo /Ox /W3 /GL /DNDEBUG /MD -Ic:\users\pichau\pycharmprojects\tccfwork\venv\lib\site-packages \numpy\core\include -Ic:\users\pichau\pycharmprojects\tccfwork\venv\lib\site-packages\numpy\core\include -Ic:\users
pichau\pycharmprojects\tccfwork\venv\include -IC:\Users\Pichau\AppData\Local\Programs\Python\Python36-32\include -IC :\Users\Pichau\AppData\Local\Programs\Python\Python36-32\include -IC:\Program Files (x86)\Microsoft Visual Studio\20 19\Community\VC\Tools\MSVC\14.27.29110\include /EHsc /Tpsrc\skmultiflow\lazy\src\libNearestNeighbors\nearestNeighbor .cpp /Fobuild\temp.win32-3.6\Release\src\skmultiflow\lazy\src\libNearestNeighbors\nearestNeighbor.obj /Zm1000" faile d with exit status 2 ---------------------------------------- ERROR: Command errored out with exit status 1: 'c:\users\pichau\pycharmprojects\tccfwork\venv\scripts\python.exe' -u -c 'import io, os, sys, setuptools, tokenize; sys.argv[0] = '"'"'C:\Users\Pichau\AppData\Local\Temp\pip-insta ll-wxbtl444\scikit-multiflow_003d6117f70b419bb0cf4433d5182b7c\setup.py'"'"'; file='"'"'C:\Users\Pichau\AppD ata\Local\Temp\pip-install-wxbtl444\scikit-multiflow_003d6117f70b419bb0cf4433d5182b7c\setup.py'"'"';f = getattr (tokenize, '"'"'open'"'"', open)(file) if os.path.exists(file) else io.StringIO('"'"'from setuptools import setup; setup()'"'"');code = f.read().replace('"'"'\r\n'"'"', '"'"'\n'"'"');f.close();exec(compile(code, file, '" '"'exec'"'"'))' install --record 'C:\Users\Pichau\AppData\Local\Temp\pip-record-tsuf4uvp\install-record.txt' --singl e-version-externally-managed --compile --install-headers 'c:\users\pichau\pycharmprojects\tccfwork\venv\include\site \python3.6\scikit-multiflow' Check the logs for full command output.

Describe the bug I have a sample stream containing durations (in seconds), and I have analyzed the change points using the ADWINChangeDetector class from MOA. The reported change points are 352, 480. When I apply the same analysis using ADWIN class from scikit-multiflow I have got different points: 416, 448, 480.

To Reproduce I have simulated the problem by comparing the results obtained from MOA and the python library. I have attached the sample file I have applied (the same data is inputted from both) and a sample code.

ADWINPython_Simulate.zip MOA_Simulate.zip

Expected behavior I expected the same chance points (352, 480).

Screenshots If applicable, add screenshots to help explain your problem.

Additional context I have already identified the difference in the code from scikit-multiflow and I am creating a pull request for it. There are two problems:

1. The method detected_change is not iterating over the complete bucket because of line 317: for k in range(cursor.bucket_size_row - 1):

It should work as the Java implementation, to fix: for k in range(cursor.bucket_size_row):

2. After detecting a change point, the detector should be reset. For fixing this I have added the following lines at the beginning of the add_element method: if self.in_concept_change: self.reset()