Mailman 3 January 2021 - scikit-learn

[Data Umbrella AFME sprint] share
by Reshama Shaikh Jan. 14, 2021

Jan. 14, 2021

Hello, There is an upcoming scikit-learn open source sprint to increase participation of folks in the **Africa and Middle East** regions. If you are located in **Africa and Middle East**, or have contacts there, please share: Data Umbrella has organized a scikit-learn open source sprint for 06-Feb-2021, with a focus on **Africa and Middle East** regions. A sprint is a 4-hour online hackathon where data scientists / developers will work with a pair programming partner on a beginner-friendly issue in the scikit-learn repo. Some knowledge of python, scikit-learn and machine learning is required. This sprint is an excellent opportunity to increase machine learning and python skills, get mentorship from core developers of the library and get started in contributing to open source. Full details are available here: https://afme2021.dataumbrella.org Also, here are social media links that can be shared: - Twitter [a] - LinkedIn [b] - Facebook [c] [a] https://twitter.com/DataUmbrella/status/1346486322958131202 [b] https://www.linkedin.com/feed/update/urn:li:activity:6752255120714579968/ [c] https://www.facebook.com/data.umbrella.dei/photos/a.156775909179975/4325969… Application deadline: 22-January-2021 We are happy to answer any questions. They can be sent to: data.umbrella.dei(a)gmail.com Best, Reshama --- Reshama Shaikh she/her Blog <https://reshamas.github.io> | Twitter <https://twitter.com/reshamas> | LinkedIn <https://www.linkedin.com/in/reshamas/> | GitHub <https://github.com/reshamas> Data Umbrella <https://www.dataumbrella.org> NYC PyLadies <https://meet.meetup.com/wf/click?upn=pEEcc35imY7Cq0tG1vyTt6zEs68RbcMfjPcajN…>

1 0

2 million samples dataset caused python and OS crash
by Liu James Jan. 8, 2021

Jan. 8, 2021

Hi all, I'm using a medium dataset KDD99 IDS( https://www.ll.mit.edu/r-d/datasets/1999-darpa-intrusion-detection-evaluati…) for model training, and the dataset has 2 million samples. When using fit_transform(), the OS crashed with log "Process 13851(python) of user xxx dumped core. Stack trace .../numpy/core/_multiarray_umath_cpython_36m_x86_64... ". The hardware: Centos 8, Intel i9, 128GB RAM, stack size is set unlimited. Such crash can be reproduced. Thanks.

3 4

extraction of grid search values
by Glenn Schultz Jan. 5, 2021

Jan. 5, 2021

All, I have a grid search of gradient boosting classifier. All works well the best model is extracted and predict works on the model. I would like to extract the cv_results_ My set-up is pretty standard gbclassifier = GridSearchCV(GradientBoostingClassifier(), parameters, verbose = 5, n_jobs = 5, cv = ShuffleSplit(n_splits = 5, test_size = .2, random_state = 42), refit = True, scoring = “roc_auc”) print(gbclassifier.cv_results_) returns an attribute error ‘Gradient Boosting Classifier’ has no attribute cv_results. I am not sure what I am doing wrong I checked the documentation and followed some SO examples but no progress. I am missing something any help is appreciated. Best, Glenn

2 1

Comparing Scikit and Xlstat for PCA analysis
by Mahmood Naderan Jan. 5, 2021

Jan. 5, 2021

Hi I have a test csv file and I have written a code to show the PCA for that. I also use another tool in Excel (XLSTAT) to compare the results. The XLSTAT automatically calculates the number of features, however, based on my understanding, I have to specify how many components are needed using the scikit package. For example, while XLSTAT shows 5 features: Factor scores: F1 F2 F3 F4 F5 A1 -1.293 -0.663 -0.462 -0.713 0.010 A2 -0.297 0.293 -1.429 0.397 0.056 A3 2.328 0.069 0.987 -0.108 0.062 A4 -0.556 -2.273 0.538 0.344 -0.032 A5 1.823 0.775 -0.597 -0.052 -0.085 A6 -2.005 1.799 0.963 0.133 -0.011 In the following code, I specified 2 components: x = StandardScaler().fit_transform(x) pca = PCA(n_components=2) principalComponents = pca.fit_transform(x) print( principalComponents ) [[-1.29292842 0.66325508] [-0.29706395 -0.29346337] [ 2.32751305 -0.06850045] [-0.5558091 2.27288988] [ 1.82312052 -0.77527304] [-2.0048321 -1.7989081 ]] As you can see, the first column in XLSTAT and scikit are the same. However, the second columns are negated. For example, considering F1 and F2, we see XLSTAT => -1.293 -0.663 scikit => [-1.29292842 0.66325508] So, my questions are 1) Isn't there any way to use scikit for an unknown number of principal components? So that I can query the number of principal components and use a scree plot then. 2) Considering the F1 and F2 as a XY scatter point, I want to know why the value of Y in XLSTAT and scikit are opposite? The code which I write is available at https://pastebin.com/ghJQ6L4C Any idea? Regards, Mahmood

2 3

IterativeImputer
by Sole Galli Jan. 4, 2021

Jan. 4, 2021

Hello team, I am reading in some of the MICE original articles that supposedly, each variable should be modelled upon the other ones in the data, with a suitable model. So for example, if the variable with NA is binary, it should be modelled with classification, or if continuous with a regression model. Am I correct to understand that this is not possible yet with the IterativeImputer? because I should set the estimator in the estimator parameter and that will be used for all variables. Is there a workaround? Thanks a lot! Regards Soledad Galli https://www.trainindata.com/

1 0