from typing import List
import numpy as np
import pandas as pd
from toolz import curry, merge, assoc
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.linear_model import LogisticRegression
from sklearn import __version__ as sk_version
from fklearn.types import LearnerReturnType, LogType
from fklearn.common_docstrings import learner_return_docstring, learner_pred_fn_docstring
from fklearn.training.utils import log_learner_time
[docs]@curry
@log_learner_time(learner_name='logistic_classification_learner')
def logistic_classification_learner(df: pd.DataFrame,
features: List[str],
target: str,
params: LogType = None,
prediction_column: str = "prediction",
weight_column: str = None) -> LearnerReturnType:
"""
Fits an logistic regression classifier to the dataset. Return the predict function
for the model and the predictions for the input dataset.
Parameters
----------
df : pandas.DataFrame
A Pandas' DataFrame with features and target columns.
The model will be trained to predict the target column
from the features.
features : list of str
A list os column names that are used as features for the model. All this names
should be in `df`.
target : str
The name of the column in `df` that should be used as target for the model.
This column should be binary, since this is a classification model.
params : dict
The LogisticRegression parameters in the format {"par_name": param}. See:
http://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html
prediction_column : str
The name of the column with the predictions from the model.
If a multiclass problem, additional prediction_column_i columns will be added for i in range(0,n_classes).
weight_column : str, optional
The name of the column with scores to weight the data.
"""
def_params = {"C": 0.1, "multi_class": "ovr"}
merged_params = def_params if not params else merge(def_params, params)
weights = df[weight_column].values if weight_column else None
clf = LogisticRegression(**merged_params)
clf.fit(df[features].values, df[target].values, sample_weight=weights)
def p(new_df: pd.DataFrame) -> pd.DataFrame:
pred = clf.predict_proba(new_df[features].values)
if merged_params["multi_class"] == "multinomial":
col_dict = {prediction_column + "_" + str(key): value for (key, value) in enumerate(pred.T)}
col_dict.update({prediction_column: pred.argmax(axis=1)})
else:
col_dict = {prediction_column: pred[:, 1]}
return new_df.assign(**col_dict)
p.__doc__ = learner_pred_fn_docstring("logistic_classification_learner")
log = {'logistic_classification_learner': {
'features': features,
'target': target,
'parameters': merged_params,
'prediction_column': prediction_column,
'package': "sklearn",
'package_version': sk_version,
'feature_importance': dict(zip(features, clf.coef_.flatten())),
'training_samples': len(df)
}}
return p, p(df), log
logistic_classification_learner.__doc__ += learner_return_docstring("Logistic Regression")
[docs]@curry
@log_learner_time(learner_name='xgb_classification_learner')
def xgb_classification_learner(df: pd.DataFrame,
features: List[str],
target: str,
learning_rate: float = 0.1,
num_estimators: int = 100,
extra_params: LogType = None,
prediction_column: str = "prediction",
weight_column: str = None) -> LearnerReturnType:
"""
Fits an XGBoost classifier to the dataset. It first generates a DMatrix
with the specified features and labels from `df`. Then, it fits a XGBoost
model to this DMatrix. Return the predict function for the model and the
predictions for the input dataset.
Parameters
----------
df : pandas.DataFrame
A Pandas' DataFrame with features and target columns.
The model will be trained to predict the target column
from the features.
features : list of str
A list os column names that are used as features for the model. All this names
should be in `df`.
target : str
The name of the column in `df` that should be used as target for the model.
This column should be binary, since this is a classification model.
learning_rate : float
Float in the range (0, 1]
Step size shrinkage used in update to prevents overfitting. After each boosting step,
we can directly get the weights of new features. and eta actually shrinks the
feature weights to make the boosting process more conservative.
See the eta hyper-parameter in:
http://xgboost.readthedocs.io/en/latest/parameter.html
num_estimators : int
Int in the range (0, inf)
Number of boosted trees to fit.
See the n_estimators hyper-parameter in:
http://xgboost.readthedocs.io/en/latest/python/python_api.html
extra_params : dict, optional
Dictionary in the format {"hyperparameter_name" : hyperparameter_value}.
Other parameters for the XGBoost model. See the list in:
http://xgboost.readthedocs.io/en/latest/parameter.html
If not passed, the default will be used.
prediction_column : str
The name of the column with the predictions from the model.
If a multiclass problem, additional prediction_column_i columns will be added for i in range(0,n_classes).
weight_column : str, optional
The name of the column with scores to weight the data.
"""
import xgboost as xgb
params = extra_params if extra_params else {}
params = assoc(params, "eta", learning_rate)
params = params if "objective" in params else assoc(params, "objective", 'binary:logistic')
weights = df[weight_column].values if weight_column else None
dtrain = xgb.DMatrix(df[features].values, label=df[target].values, feature_names=map(str, features), weight=weights)
bst = xgb.train(params, dtrain, num_estimators)
def p(new_df: pd.DataFrame, apply_shap: bool = False) -> pd.DataFrame:
dtest = xgb.DMatrix(new_df[features].values, feature_names=map(str, features))
pred = bst.predict(dtest)
if params["objective"] == "multi:softprob":
col_dict = {prediction_column + "_" + str(key): value
for (key, value) in enumerate(pred.T)}
col_dict.update({prediction_column: pred.argmax(axis=1)})
else:
col_dict = {prediction_column: pred}
if apply_shap:
import shap
explainer = shap.TreeExplainer(bst)
shap_values = list(explainer.shap_values(new_df[features]))
shap_expected_value = explainer.expected_value
shap_output = {"shap_values": shap_values,
"shap_expected_value": np.repeat(shap_expected_value, len(shap_values))}
if params["objective"] == "multi:softprob":
raise NotImplementedError("SHAP values are not implemented for multiclass XGBoost in fklearn")
else:
col_dict = merge(col_dict, shap_output)
return new_df.assign(**col_dict)
p.__doc__ = learner_pred_fn_docstring("xgb_classification_learner", shap=True)
log = {'xgb_classification_learner': {
'features': features,
'target': target,
'prediction_column': prediction_column,
'package': "xgboost",
'package_version': xgb.__version__,
'parameters': assoc(params, "num_estimators", num_estimators),
'feature_importance': bst.get_score(),
'training_samples': len(df)
}}
return p, p(df), log
xgb_classification_learner.__doc__ += learner_return_docstring("XGboost Classifier")
[docs]@curry
@log_learner_time(learner_name='nlp_logistic_classification_learner')
def nlp_logistic_classification_learner(df: pd.DataFrame,
text_feature_cols: List[str],
target: str,
vectorizer_params: LogType = None,
logistic_params: LogType = None,
prediction_column: str = "prediction") -> LearnerReturnType:
"""
Fits a text vectorizer (TfidfVectorizer) followed by
a logistic regression (LogisticRegression).
Parameters
----------
df : pandas.DataFrame
A Pandas' DataFrame with features and target columns.
The model will be trained to predict the target column
from the features.
text_feature_cols : list of str
A list of column names of the text features used for the model. All these names
should be in `df`.
target : str
The name of the column in `df` that should be used as target for the model.
This column should be binary, since this is a classification model.
vectorizer_params : dict
The TfidfVectorizer parameters in the format {"par_name": param}. See:
http://scikit-learn.org/stable/modules/generated/sklearn.feature_extraction.text.TfidfVectorizer.html
logistic_params : dict
The LogisticRegression parameters in the format {"par_name": param}. See:
http://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html
prediction_column : str
The name of the column with the predictions from the model.
"""
# set default params
default_vect_params = {"strip_accents": "unicode", "min_df": 20}
merged_vect_params = default_vect_params if not vectorizer_params else merge(default_vect_params, vectorizer_params)
default_clf_params = {"C": 0.1, "multi_class": "ovr"}
merged_logistic_params = default_clf_params if not logistic_params else merge(default_clf_params, logistic_params)
vect = TfidfVectorizer(**merged_vect_params)
clf = LogisticRegression(**merged_logistic_params)
text_df = df[text_feature_cols].apply(lambda x: x + " ", axis=1).sum(axis=1)
vect.fit(text_df.values)
sparse_vect = vect.transform(text_df.values)
clf.fit(sparse_vect, df[target].values)
def p(new_df: pd.DataFrame) -> pd.DataFrame:
predict_text_df = new_df[text_feature_cols].apply(lambda x: x + " ", axis=1).sum(axis=1)
predict_sparse_vect = vect.transform(predict_text_df)
if merged_logistic_params["multi_class"] == "multinomial":
col_dict = {prediction_column + "_" + str(key): value
for (key, value) in enumerate(clf.predict_proba(predict_sparse_vect).T)}
else:
col_dict = {prediction_column: clf.predict_proba(predict_sparse_vect)[:, 1]}
return new_df.assign(**col_dict)
p.__doc__ = learner_pred_fn_docstring("nlp_logistic_classification_learner")
params = {"vectorizer_params": merged_vect_params,
"logistic_params": merged_logistic_params}
log = {'nlp_logistic_classification_learner': {
'features': text_feature_cols,
'target': target,
'prediction_column': prediction_column,
'parameters': assoc(params, "vocab_size", sparse_vect.shape[1]),
'package': "sklearn",
'package_version': sk_version,
'training_samples': len(df)
}}
return p, p(df), log
nlp_logistic_classification_learner.__doc__ += learner_return_docstring("NLP Logistic Regression")
[docs]@curry
@log_learner_time(learner_name='lgbm_classification_learner')
def lgbm_classification_learner(df: pd.DataFrame,
features: List[str],
target: str,
learning_rate: float = 0.1,
num_estimators: int = 100,
extra_params: LogType = None,
prediction_column: str = "prediction",
weight_column: str = None) -> LearnerReturnType:
"""
Fits an LGBM classifier to the dataset.
It first generates a Dataset
with the specified features and labels from `df`. Then, it fits a LGBM
model to this Dataset. Return the predict function for the model and the
predictions for the input dataset.
Parameters
----------
df : pandas.DataFrame
A pandas DataFrame with features and target columns.
The model will be trained to predict the target column
from the features.
features : list of str
A list os column names that are used as features for the model. All this names
should be in `df`.
target : str
The name of the column in `df` that should be used as target for the model.
This column should be binary, since this is a classification model.
learning_rate : float
Float in the range (0, 1]
Step size shrinkage used in update to prevents overfitting. After each boosting step,
we can directly get the weights of new features. and eta actually shrinks the
feature weights to make the boosting process more conservative.
See the learning_rate hyper-parameter in:
https://github.com/Microsoft/LightGBM/blob/master/docs/Parameters.rst
num_estimators : int
Int in the range (0, inf)
Number of boosted trees to fit.
See the num_iterations hyper-parameter in:
https://github.com/Microsoft/LightGBM/blob/master/docs/Parameters.rst
extra_params : dict, optional
Dictionary in the format {"hyperparameter_name" : hyperparameter_value}.
Other parameters for the LGBM model. See the list in:
https://github.com/Microsoft/LightGBM/blob/master/docs/Parameters.rst
If not passed, the default will be used.
prediction_column : str
The name of the column with the predictions from the model.
weight_column : str, optional
The name of the column with scores to weight the data.
"""
import lightgbm as lgbm
params = extra_params if extra_params else {}
params = assoc(params, "eta", learning_rate)
params = params if "objective" in params else assoc(params, "objective", 'binary')
weights = df[weight_column].values if weight_column else None
dtrain = lgbm.Dataset(df[features].values, label=df[target], feature_name=list(map(str, features)), weight=weights,
silent=True)
bst = lgbm.train(params, dtrain, num_estimators)
def p(new_df: pd.DataFrame, apply_shap: bool = False) -> pd.DataFrame:
if params["objective"] == "multiclass":
col_dict = {prediction_column + "_" + str(key): value
for (key, value) in enumerate(bst.predict(new_df[features].values).T)}
else:
col_dict = {prediction_column: bst.predict(new_df[features].values)}
if apply_shap:
import shap
explainer = shap.TreeExplainer(bst)
shap_values = list(explainer.shap_values(new_df[features]))
shap_expected_value = explainer.expected_value
shap_output = {"shap_values": shap_values,
"shap_expected_value": np.repeat(shap_expected_value, len(shap_values))}
if params["objective"] == "multiclass":
raise NotImplementedError("SHAP values are not implemented for multiclass LGBM in fkit-learn")
else:
col_dict = merge(col_dict, shap_output)
return new_df.assign(**col_dict)
p.__doc__ = learner_pred_fn_docstring("lgbm_classification_learner", shap=True)
log = {'lgbm_classification_learner': {
'features': features,
'target': target,
'prediction_column': prediction_column,
'package': "lightgbm",
'package_version': lgbm.__version__,
'parameters': assoc(params, "num_estimators", num_estimators),
'feature_importance': dict(zip(features, bst.feature_importance().tolist())),
'training_samples': len(df)},
'object': bst}
return p, p(df), log
lgbm_classification_learner.__doc__ += learner_return_docstring("LGBM Classifier")