ブループリントワークショップの基本ステップノートブック¶

Binning of numerical variables (BINNING)

  Invalid value(s) supplied
    max_bins (b) = -22
      - Must be a 'intgrid' parameter defined by: [2, 500]

Failed to save: parameter validation failed.

Keras Neural Network Classifier (KERASC)

All parameters valid!

Help on PNI2 in module datarobot_bp_workshop.factories object:

class PNI2(datarobot_bp_workshop.friendly_repr.FriendlyRepr)
 |  Missing Values Imputed (quick median)
 |  
 |  Impute missing values on numeric variables with their median and create indicator variables to mark imputed records 
 |  
 |  Parameters
 |  ----------
 |  output_method: string, one of (TaskOutputMethod.TRANSFORM).
 |  task_parameters: dict, which may contain:
 |  
 |    scale_small (s): select, (Default=0)
 |      Possible Values: [False, True]
 |  
 |    threshold (t): int, (Default=10)
 |      Possible Values: [1, 99999]
 |  
 |  Method resolution order:
 |      PNI2
 |      datarobot_bp_workshop.friendly_repr.FriendlyRepr
 |      builtins.object
 |  
 |  Methods defined here:
 |  
 |  __call__(zelf, *inputs, output_method=None, task_parameters=None, output_method_parameters=None, x_transformations=None, y_transformations=None, freeze=False, version=None)
 |  
 |  __friendly_repr__(zelf)
 |  
 |  documentation(zelf, auto_open=False)
 |  
 |  ----------------------------------------------------------------------
 |  Data and other attributes defined here:
 |  
 |  description = 'Impute missing values on numeric variables with ...eate...
 |  
 |  label = 'Missing Values Imputed (quick median)'
 |  
 |  task_code = 'PNI2'
 |  
 |  task_parameters = scale_small (s): select, (Default=0)
 |  
 |  threshold (t):...
 |  
 |  ----------------------------------------------------------------------
 |  Methods inherited from datarobot_bp_workshop.friendly_repr.FriendlyRepr:
 |  
 |  __repr__(self)
 |      Return repr(self).
 |  
 |  ----------------------------------------------------------------------
 |  Data descriptors inherited from datarobot_bp_workshop.friendly_repr.FriendlyRepr:
 |  
 |  __dict__
 |      dictionary for instance variables (if defined)
 |  
 |  __weakref__
 |      list of weak references to the object (if defined)

Custom

  - Awesome Model (CUSTOMR_6019ae978cc598a46199cee1)
  - "My Custom Task" (CUSTOMR_608e42ac186a7242380a6a98)
  - "My Custom Task" (CUSTOMR_608e42ecd5eb0dc5f28d0dda)
  - "My Custom Task" (CUSTOMR_608e43fc01f9f466aa8d0d81)
  - My Custom Ridge Regressor w/ Imputation (CUSTOMR_608e5a4ed5eb0dc5f28d0ea0)
  - My Custom Ridge Regressor w/ Imputation (CUSTOMR_608e5bc8b66a4934d58d0d4e)
  - My Custom Ridge Regressor w/ Imputation (CUSTOMR_608ef72b6f13f54305667783)
  - My Custom Ridge Regressor w/ Imputation (CUSTOMR_608ef74c5dda651931052422)
  - Second model (CUSTOMC_6019d18adfa83afbad99cdb8)
  - My Imputation Task (CUSTOMT_6188b0e6fb465717f029fd05)
  - Image Featurizer (CUSTOMT_61b452e57fd5b0629a2f4fd3)
  - Maybe Broken? (CUSTOMT_61b7d3f26f8e01a1a8f7bc0c)
Preprocessing

  Numeric Preprocessing

    Data Quality

      - Numeric Data Cleansing (NDC)
    Dimensionality Reducer

      - Truncated Singular Value Decomposition (SVD2)
      - Partial Principal Components Analysis (PPCA)
      - Truncated Singular Value Decomposition (SVD)
    Scaling

      - Impose Uniform Transform (UNIF3)
      - Log Transformer (LOGT)
      - Smooth Ridit Transform (RDT5)
      - Standardize (RST)
      - Search for best transformation including Smooth Ridit (BTRANSF6)
      - Transparent Search for best transformation (BTRANSF6T)
      - Transform on the link function scale (LINK)
      - Ridit Transform (SRDT3)
      - Standardize (ST)
    - Sparse Interaction Machine (SPOLY)
    - Constant Splines (GS)
    - One-Hot Encoding (PDM3)
    - Numeric Data Cleansing (NDC)
    - Missing Values Imputed (quick median) (PNI2)
    - Missing Values Imputed (arbitrary or quick median) (PNIA4)
    - Normalizer (NORM)
    - Search for ratios (RATIO3)
    - Binning of numerical variables (BINNING)
    - Search for differences (DIFF3)
  Categorical Preprocessing

    - Categorical Embedding (CATEMB)
    - Category Count (PCCAT)
    - One-Hot Encoding (PDM3)
    - Ordinal encoding of categorical variables (ORDCAT2)
    - Univariate credibility estimates with L2 (CRED1b1)
    - Buhlmann credibility estimates for high cardinality features (CRED1)
  Text Preprocessing

    - TextBlob Sentiment Featurizer (TEXTBLOB_SENTIMENT)
    - NLTK Sentiment Featurizer (NLTK_SENTIMENT)
    - One-Hot Encoding (PDM3)
    - Pretrained TinyBERT Featurizer (TINYBERTFEA)
    - SpaCy Named Entity Recognition Detector (SPACY_NAMED_ENTITY_RECOGNITION)
    - Fasttext Word Vectorization and Mean text embedding (TXTEM1)
    - Keras encoding of text variables (KERAS_TOKENIZER)
    - Matrix of word-grams occurrences (PTM3)
  Image Preprocessing

    - OpenCV Detect Largest Rectangle (OPENCV_DETECT_LARGEST_RECTANGLE)
    - OpenCV Image Featurizer (OPENCV_FEATURIZER)
    - Grayscale Downscaled Image Featurizer (IMG_GRAYSCALE_DOWNSCALED_IMAGE_FEATURIZER)
    - No Post Processing (IMAGE_POST_PROCESSOR)
    - Pretrained Multi-Level Global Average Pooling Image Featurizer (IMGFEA)
  Summarized Categorical Preprocessing

    - Summarized Categorical to Sparse Matrix (CDICT2SP)
    - Single Column Converter for Summarized Categorical (SCBAGOFCAT2)
  Geospatial Preprocessing

    - Spatial Neighborhood Featurizer (GEO_NEIGHBOR_V1)
    - Geospatial Location Converter (GEO_IN)

Models

  Regression

    - eXtreme Gradient Boosted Trees Quantile Regressor with Early Stopping (ESQUANTXGBR)
    - ExtraTrees Regressor (RFR)
    - Elastic-Net Regressor (L1 / Least-Squares Loss) (ENETCDWC)
    - Light Gradient Boosted Trees Regressor with Early Stopping (ESLGBMTR)
    - eXtreme Gradient Boosted Trees Regressor (PXGBR2)
    - Ridge Regression (RIDGE)
    - Nystroem Kernel SVM Regressor (ASVMER)
    - eXtreme Gradient Boosted Trees Regressor with Early Stopping and Unsupervised Learning Features (UESXGBR2)
    - eXtreme Gradient Boosted Trees Regressor (XGBR2)
    - eXtreme Gradient Boosted Trees Regressor (XL_PXGBR2)
    - Nystroem Kernel SVM Regressor (ASVMSKR)
    - Partial Least-Squares Regression (PLS)
    - Gaussian Process Regressor with Rational Quadratic Kernel (GPRRQ)
    - Eureqa Regressor (EQR)
    - Auto-Tuned Char N-Gram Text Modeler using token counts (CNGER2)
    - Frequency-Severity Generalized Additive Model (FSGG2)
    - Hot Spots (XPRIMR)
    - Linear Regression (GLMCD)
    - Frequency-Severity ElasticNet (FSEE)
    - Gradient Boosted Trees Regressor with Early Stopping (Least-Squares Loss) (ESGBR2)
    - Light Gradient Boosting on ElasticNet Predictions (RES_ESLGBMTR)
    - Support Vector Regressor (Radial Kernel) (SVMR2)
    - Regularized Quantile Regressor with Keras (KERAS_REGULARIZED_QUANTILE_REG)
    - Auto-tuned K-Nearest Neighbors Regressor (Euclidean Distance) (KNNR)
    - Lasso Regression (LASSO2)
    - Gaussian Process Regressor with Radial Basis Function Kernel (GPRRBF)
    - XRuleFit Regressor (XRULEFITR)
    - Frequency-Severity Light Gradient Boosted Trees (FSLL)
    - Ridge Regression (RIDGEWC)
    - Stochastic Gradient Descent Regression (SGDR)
    - Eureqa Generalized Additive Model (EQ_ESXGBR)
    - Elastic-Net Regressor (L1 / Least-Squares Loss) with K-Means Distance Features (KMDENETCD)
    - eXtreme Gradient Boosting on ElasticNet Predictions (RES_XGBR2)
    - Auto-Tuned Word N-Gram Text Modeler using token counts (WNGER2)
    - Auto-Tuned Summarized Categorical Modeler (SCENETR)
    - Keras Neural Network Regressor (KERASR)
    - Elastic-Net Regressor (L1 / Least-Squares Loss) (ENETCD)
    - eXtreme Gradient Boosted Trees Regressor (XL_XGBR2)
    - Gaussian Process Regressor with Dot Product Kernel (GPRDP)
    - Dropout Additive Regression Trees Regressor (PLGBMDR)
    - Elastic-Net Regressor (L1 / Least-Squares Loss) with Binned numeric features (BENETCD2)
    - eXtreme Gradient Boosted Trees Regressor with Early Stopping (XL_ESXGBR2)
    - Auto-tuned Stochastic Gradient Descent Regression (SGDRA)
    - RuleFit Regressor (RULEFITR)
    - Gaussian Process Regressor with Exponential Sine Squared Kernel (GPRESS)
    - Adaboost Regressor (ABR)
    - Elastic-Net Regressor (L1 / Least-Squares Loss) with Unsupervised Learning Features (UENETCD)
    - Gaussian Process Regressor with Matern Kernel (GPRM)
    - Light Gradient Boosting on ElasticNet Predictions (RES_PLGBMTR)
    - Gradient Boosted Trees Quantile Regressor with Early Stopping (QESGBR2)
    - ExtraTrees Regressor (Shallow) (SHAPRFR)
    - Statsmodels Quantile Regressor (QUANTILER)
    - eXtreme Gradient Boosted Trees Regressor with Early Stopping (ESXGBR2)
    - LightGBM Random Forest Regressor (PLGBMRFR)
    - Frequency-Cost ElasticNet (FCEE)
    - Frequency-Severity eXtreme Gradient Boosted Trees (FSXX2)
    - Gradient Boosted Trees Quantile Regressor (QGBR2)
    - eXtreme Gradient Boosting on ElasticNet Predictions (RES_ESXGBR2)
  Binary Classification

    - Stochastic Gradient Descent Classifier (SGDC)
    - LightGBM Random Forest Classifier (PLGBMRFC)
    - Bernoulli Naive Bayes classifier (scikit-learn) (BNBC)
    - Dropout Additive Regression Trees Classifier (PLGBMDC)
    - Auto-Tuned Char N-Gram Text Modeler using token counts (CNGEC2)
    - Gaussian Process Classifier with Matern Kernel (GPCM)
    - Gradient Boosted Trees Classifier with Early Stopping (ESGBC)
    - XRuleFit Classifier (XRULEFITC)
    - Support Vector Classifier (Radial Kernel) (SVMC2)
    - Multinomial Naive Bayes classifier (scikit-learn) (MNBC)
    - Adaboost Classifier (ABC)
    - eXtreme Gradient Boosting on ElasticNet Predictions (RES_XGBC2)
    - Elastic-Net Classifier (L1 / Binomial Deviance) (LENETCDWC)
    - Gaussian Process Classifier with Radial Basis Function Kernel (GPCRBF)
    - Light Gradient Boosted Trees Classifier with Early Stopping (ESLGBMTC)
    - Eureqa Classifier (EQC)
    - ExtraTrees Classifier (Gini) (SHAPRFC)
    - Logistic Regression (LR)
    - Keras Neural Network Classifier (KERASC)
    - Nystroem Kernel SVM Classifier (ASVMEC)
    - eXtreme Gradient Boosted Trees Classifier with Early Stopping and Unsupervised Learning Features (UESXGBC2)
    - RuleFit Classifier (RULEFITC)
    - Regularized Logistic Regression (L2) (LR1)
    - Nystroem Kernel SVM Classifier (ASVMSKC)
    - Naive Bayes combiner classifier (CNBC)
    - Light Gradient Boosted Trees Classifier with Early Stopping and Unsupervised Learning Features (UESLGBMTC)
    - Light Gradient Boosting on ElasticNet Predictions (RES_PLGBMTC)
    - Elastic-Net Classifier (L1 / Binomial Deviance) with K-Means Distance Features (KMDLENETCD)
    - ExtraTrees Classifier (Gini) (RFC)
    - Hot Spots (XPRIMC)
    - Partial Least-Squares Classification (PLSC)
    - Auto-tuned K-Nearest Neighbors Classifier (Euclidean Distance) (KNNC)
    - Eureqa Generalized Additive Model Classifier (EQ_ESXGBC)
    - eXtreme Gradient Boosted Trees Classifier (XL_XGBC2)
    - Auto-Tuned Summarized Categorical Modeler (SCLENETC)
    - Light Gradient Boosting on ElasticNet Predictions (RES_ESLGBMTC)
    - Elastic-Net Classifier with Naive Bayes Feature Weighting (NB_LENETCD)
    - eXtreme Gradient Boosted Trees Classifier (XGBC2)
    - Elastic-Net Classifier (L1 / Binomial Deviance) (LENETCD)
    - Gaussian Naive Bayes classifier (scikit-learn) (GNBC)
    - Logistic Regression (LRCD)
    - eXtreme Gradient Boosted Trees Classifier with Early Stopping (ESXGBC2)
    - eXtreme Gradient Boosted Trees Classifier (PXGBC2)
    - Auto-Tuned Word N-Gram Text Modeler using token counts (WNGEC2)
  Multi-class Classification

    - Stochastic Gradient Descent Classifier (SGDC)
    - LightGBM Random Forest Classifier (PLGBMRFC)
    - Dropout Additive Regression Trees Classifier (PLGBMDC)
    - Gradient Boosted Trees Classifier with Early Stopping (ESGBC)
    - Light Gradient Boosted Trees Classifier with Early Stopping (ESLGBMTC)
    - ExtraTrees Classifier (Gini) (SHAPRFC)
    - Logistic Regression (LR)
    - Regularized Logistic Regression (L2) (LR1)
    - Light Gradient Boosted Trees Classifier with Early Stopping and Unsupervised Learning Features (UESLGBMTC)
    - Light Gradient Boosting on ElasticNet Predictions (RES_PLGBMTC)
    - Keras Neural Network Classifier (KERASMULTIC)
    - ExtraTrees Classifier (Gini) (RFC)
    - Light Gradient Boosting on ElasticNet Predictions (RES_ESLGBMTC)
    - eXtreme Gradient Boosted Trees Classifier (XGBC2)
    - Elastic-Net Classifier (L1 / Binomial Deviance) (LENETCD)
    - Logistic Regression (LRCD)
    - eXtreme Gradient Boosted Trees Classifier with Early Stopping (ESXGBC2)
    - eXtreme Gradient Boosted Trees Classifier (PXGBC2)
  Boosting

    - eXtreme Gradient Boosted Trees Regressor (XL_PXGBR2)
    - eXtreme Gradient Boosting on ElasticNet Predictions (RES_XGBC2)
    - Light Gradient Boosting on ElasticNet Predictions (RES_ESLGBMTR)
    - eXtreme Gradient Boosting on ElasticNet Predictions (RES_XGBR2)
    - eXtreme Gradient Boosted Trees Regressor (XL_XGBR2)
    - Light Gradient Boosting on ElasticNet Predictions (RES_PLGBMTC)
    - eXtreme Gradient Boosted Trees Regressor with Early Stopping (XL_ESXGBR2)
    - eXtreme Gradient Boosted Trees Classifier (XL_XGBC2)
    - Light Gradient Boosting on ElasticNet Predictions (RES_ESLGBMTC)
    - Light Gradient Boosting on ElasticNet Predictions (RES_PLGBMTR)
    - eXtreme Gradient Boosting on ElasticNet Predictions (RES_ESXGBR2)
  Unsupervised

    Anomaly Detection

      - Local Outlier Factor Anomaly Detection (ADLOF)
      - Mahalanobis Distance Ranked Anomaly Detection with PCA and Calibration (ADMAHAL_PCA_CAL)
      - Keras Autoencoder (KERAS_AUTOENCODER)
      - Isolation Forest Anomaly Detection (ADISOFOR)
      - Mahalanobis Distance Ranked Anomaly Detection with PCA (ADMahalPCA)
      - Keras Autoencoder with Calibration (KERAS_AUTOENCODER_CAL)
      - Isolation Forest Anomaly Detection with Calibration (ADISOFOR_CAL)
      - Double Median Absolute Deviation Anomaly Detection (ADDMAD)
      - Keras Variational Autoencoder (KERAS_VARIATIONAL_AUTOENCODER)
      - Keras Variational Autoencoder with Calibration (KERAS_VARIATIONAL_AUTOENCODER_CAL)
      - One-Class SVM Anomaly Detection with Calibration (ADOSVM_CAL)
      - Local Outlier Factor Anomaly Detection with Calibration (ADLOF_CAL)
      - Anomaly Detection with Supervised Learning (XGB) (ADXGB)
      - One-Class SVM Anomaly Detection (ADOSVM)
      - Anomaly Detection with Supervised Learning (XGB) and Calibration (ADXGB2_CAL)
      - Double Median Absolute Deviation Anomaly Detection with Calibration (ADDMAD_CAL)
    Clustering

      - K-Means Clustering (KMEANS)
  

Calibration

  - Calibrate predictions with RF (CALIB_V2_RFC)
  - Text fit on Residuals (L1 /  Least-Squares Loss) (XL_ENETCD)
  - Calibrate predictions: Weighted Calibration (SWCAL)
  - Calibrate predictions (CALIB)
  - Text fit on Residuals (L1 / Binomial Deviance) (XL_LENETCD)
  - Fit High Cardinality and Text (XLF_LENETCD)
  - Text fit on Residuals (L1 /  Least-Squares Loss) (RES_FDENETCD)
  - Calibrate predictions (CALIB2)
  - Calibrate predictions: Platt (PLACAL2)
  - Fit High Cardinality and Text (XLF_ENETCD)
Other

  Column Selection

    - Converter for Text Mining (SCTXT2)
    - Single Column Converter for Summarized Categorical (SCBAGOFCAT)
    - Single Column Converter (SCPICK2)
    - Single Column Converter (SCPICK)
    - Converter for Text Mining (SCTXT4)
    - Multiple Column Selector (MCPICK)
  Automatic Feature Selection

    - Feature Selection for Ratios/Differences (FS_RFR2)
    - Feature Selection for dimensionality reduction (FS_RFCDR2)
    - Feature Selection for dimensionality reduction (FS_RFCDR_LASSO)
    - Feature Selection for dimensionality reduction (FS_RFRDR_LASSO)
    - Feature Selection using L1 Regularization (FS_XL_LASSO2)
    - Rare Feature Masking (RFMASK)
    - Feature Selection for Ratios/Differences (FS_RFC2)
    - Feature Selection for dimensionality reduction (FS_RFRDR2)
  - Bind branches (BIND)

Binning of numerical variables (BINNING)

  Invalid value(s) supplied
    max_bins (b) = -22
      - Must be a 'intgrid' parameter defined by: [2, 500]

Binning of numerical variables (BINNING)

All parameters valid!