As a user trying to run the example, I often copy and paste the code snippets. When doing this with the code snippets in 1.1, I hit an error when copying the plot_result function.

>>> def plot_result(model, data, label):
...     xx, yy = np.meshgrid(np.arange(-2,1,0.01),np.arange(-2,2,0.01))
...     xx1 = np.array(xx).flatten()
...     yy1 = np.array(yy).flatten()
...     _predict = np.array(model.predict(pd.DataFrame({"X": xx1, "Y": yy1})))
... 
>>>     plt.figure(figsize=(8,8))
  File "<stdin>", line 1
    plt.figure(figsize=(8,8))
    ^
IndentationError: unexpected indent

It looks like plot_result has a blank line that is missing a tab which causes the indentation to be lost -- where other code that does contain blank lines work fine when there is a tab.

Please see this page for reference:
https://github.com/Microsoft/NimbusML-Samples/blob/master/samples/1.1%20%5BNumeric%5D%20Classification%20with%20Synthetic%20Data.ipynb

In tutorial 1.1 in cell 7 when we call plot_result(model, dataTest, labelTest) we get the following KeyError:

---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)
<ipython-input-9-f15b92a14334> in <module>
----> 1 plot_result(model, dataTest, labelTest);

<ipython-input-2-31c7838de467> in plot_result(model, data, label)
     29     plt.figure(figsize=(8,8))
     30     plt.contourf(xx,yy,_predict.reshape((400,300)), alpha = 0.2)
---> 31     plt.scatter(data["X"], data["Y"], c = label)
     32     plt.show()

c:\users\mohoov\source\repos\nimbusml-samples\tests\dependencies\python3.6\lib\site-packages\matplotlib\pyplot.py in scatter(x, y, s, c, marker, cmap, norm, vmin, vmax, alpha, linewidths, verts, edgecolors, data, **kwargs)
   2862         vmin=vmin, vmax=vmax, alpha=alpha, linewidths=linewidths,
   2863         verts=verts, edgecolors=edgecolors, **({"data": data} if data
-> 2864         is not None else {}), **kwargs)
   2865     sci(__ret)
   2866     return __ret

c:\users\mohoov\source\repos\nimbusml-samples\tests\dependencies\python3.6\lib\site-packages\matplotlib\__init__.py in inner(ax, data, *args, **kwargs)
   1803                         "the Matplotlib list!)" % (label_namer, func.__name__),
   1804                         RuntimeWarning, stacklevel=2)
-> 1805             return func(ax, *args, **kwargs)
   1806 
   1807         inner.__doc__ = _add_data_doc(inner.__doc__,

c:\users\mohoov\source\repos\nimbusml-samples\tests\dependencies\python3.6\lib\site-packages\matplotlib\axes\_axes.py in scatter(self, x, y, s, c, marker, cmap, norm, vmin, vmax, alpha, linewidths, verts, edgecolors, **kwargs)
   4193                 isinstance(c, str) or
   4194                 (isinstance(c, collections.Iterable) and
-> 4195                     isinstance(c[0], str))):
   4196             c_array = None
   4197         else:

c:\users\mohoov\source\repos\nimbusml-samples\tests\dependencies\python3.6\lib\site-packages\pandas\core\series.py in __getitem__(self, key)
    765         key = com._apply_if_callable(key, self)
    766         try:
--> 767             result = self.index.get_value(self, key)
    768 
    769             if not is_scalar(result):

c:\users\mohoov\source\repos\nimbusml-samples\tests\dependencies\python3.6\lib\site-packages\pandas\core\indexes\base.py in get_value(self, series, key)
   3116         try:
   3117             return self._engine.get_value(s, k,
-> 3118                                           tz=getattr(series.dtype, 'tz', None))
   3119         except KeyError as e1:
   3120             if len(self) > 0 and self.inferred_type in ['integer', 'boolean']:

pandas\_libs\index.pyx in pandas._libs.index.IndexEngine.get_value()

pandas\_libs\index.pyx in pandas._libs.index.IndexEngine.get_value()

pandas\_libs\index.pyx in pandas._libs.index.IndexEngine.get_loc()

pandas\_libs\index.pyx in pandas._libs.index.IndexEngine._get_loc_duplicates()

pandas\_libs\index_class_helper.pxi in pandas._libs.index.Int64Engine._maybe_get_bool_indexer()

KeyError: 0

This appears to be a chance from Matplotlib v3.0.0 to v3.0.1 because they are doing a check for c[0] in the color parameter where we have c=labelTest, which is a pandas Series which is indexed by data point ID's and it happens to not have a zero index.

A Medium article on NimbusML would be a great way to showcase its capabilities in a succinct and appealing way. We should write a NimbusML Medium article with an easy-to-follow example. This Medium article should be authored by an official Microsoft account, or co-authored by Microsoft software engineers.

Now that there is time series support in NimbusML, should a time series sample be added?

1. Go to this repo's readme
2. Click on the azure notebooks link

Expect: opens NimbusML samples in Azure notebooks
Actual: opens Microsoft's home page.

Describe the bug
The link to documentation points to: https://docs.microsoft.com/en-us/nimbusml/tutorials, which gives a 404

To Reproduce
Steps to reproduce the behavior:

1. Go to github page, scroll to the paragraph that reads:

This is an open source project located at https://github.com/Microsoft/NimbusML. We'd love for you to try it out and/or contribute! For a full list of the samples/notebooks, please refer to our documentation.

2. Click on 'documentation'
3. See that there is a 404

Expected behavior
This should link to the live documentation site.

Describe the bug
The Image Classification example is producing different results at each run. This is due to the fact that there are non-deterministic factors involved in producing the output, such as the random initialization of the k-means centroids, seeds given to the pipeline (MLContext) itself, to the PCATransformer, etc.

To Reproduce
Run the Image Classification example (on Jupyter or local machine) multiple times, and the user will see that tables with different scores and values are displayed, which also results in the final PCA graph of the image cluster displaying different plots.

Expected behavior
Every single run of the Image Classification example should produce the exact same output. It should be deterministic.

Additional context
This was first noticed while looking at Issue 23.

Describe the bug
In the build of commit #1186afe on Azure DevOps, the job fails during the execution of tests\cmd.exe .

However, when ran locally, this build does not always fail. In fact, when ran locally on Jupyter Notebook, it builds successfully, but produces different results than the expected results that are in the GitHub file for the clustering sample file.

To Reproduce
Build and Log file: https://dev.azure.com/aifx/public/_build/results?buildId=1486

The most important lines in the log file are lines 496-581, Specifically, the failed build originates here, when the trained pipeline is called to make a prediction and generate clustering results.

Screenshots

Expected results (https://github.com/mstfbl/NimbusML-Samples/blob/master/samples/2.6%20%5BImage%5D%20Image%20Processing%20-%20Clustering.ipynb, seems to be running Python 3.7.2):

Actual results 1 (from local Jupyter Notebook, Python 3.6.6):

Actual results 2 (from local .py file, converted from .ipynb, Python 3.6.6):

Describe the bug
I was trying to merge the commit for fixing the issue of using wrong parameter names (#15 ). However I got an ImportError while merging the branch, so I change the line of installing the python package from "pip install https://pythonpkgdeps.blob.core.windows.net/pytlc/nimbusml-0.6.0-cp36-none-win_amd64.whl" to "pip install https://pythonpkgdeps.blob.core.windows.net/pytlc/nimbusml-1.0.0-cp37-none-win_amd64.whl", but it still doesn't work. I also tried to see what version is available through the URL, but the page was not showing anything.

In the Introduction section, there is a reference to an image, but no image is displayed. Instead this is shown:

<img src="images\1-3-1.png" align = "middle"/>

Please see this link for reference:
https://github.com/Microsoft/NimbusML-Samples/blob/master/samples/1.2%20%5BNumeric%5D%20Regression%20with%20Synthetic%20Data.ipynb

E.g. in this tutorial, we have the following:

In this example, we develop a similar model as in the tutorial for Twitter Data 1. This points to a 404.

The correct link is https://docs.microsoft.com/en-us/nimbusml/tutorials/b_a-sentiment-analysis-1-data-loading-with-pandas

There are other such cases as well.

Do a pass over the notebooks and fix all broken links.

Related #13