Felipe Brandão vanbrands

Staff Data Analyst @Typeform

vanbrands / mlprod:tuning.py

Last active July 30, 2020 03:40

	from sklearn.preprocessing import KBinsDiscretizer
	from sklearn.model_selection import RandomizedSearchCV

	search_space = {
	'preprocessing__binfeatures__n_bins': [10, 20],
	'model__n_estimators': [500, 1000, 2000],
	'model__class_weight': ['balanced', 'balanced_subsample']
	}

	preprocessing = ColumnTransformer(transformers=[

vanbrands / mlprod:pickle.py

Last active July 30, 2020 02:57

	# Persist pipeline to disk
	pipeline.fit(x_train, y_train)
	joblib.dump(pipeline, 'model.joblib')

	# Load pipeline and make prediction
	pipeline = joblib.load('model.joblib')
	y_hat = pipeline.predict(x)

vanbrands / mlprod:pipeline2.py

Last active July 23, 2020 17:03

	from sklearn.compose import ColumnTransformer

	# Define you transformers.
	preprocessing = ColumnTransformer(transformers=[
	('encode_categorical_features', OneHotEncoder(), ['categorical_column'])
	], remainder='passthrough')

	estimator = GradientBoostingClassifier()

	# Define steps in the pipeline.

vanbrands / mlprod:onehot.py

Last active July 24, 2020 20:11

	from sklearn.preprocessing import OneHotEncoder

	encoder = OneHotEncoder()
	dummies = encoder.fit_transform(df['categorical_column'])

vanbrands / mlprod:categorical.py

Last active July 10, 2020 01:54

	# You have you set of possible categories in a list for example.
	# (it could be anything really...)
	possible_categories = ['foo', 'bar]

	# Convert the categorical column to type Categorical
	df['categorical_column'] = pd.Categorical(
	values=df['categorical_column'],
	categories=possible_categories
	)

vanbrands / mlprod:dummies.py

Last active July 24, 2020 20:15

	import pandas as pd

	dummies = pd.get_dummies(df, columns=['categorical_colum1', 'categorical_colum2'])

vanbrands / production1.py

Created July 10, 2020 01:45

dummies = pd.get_dummies(df['categorical_column'])

vanbrands / deploy_notebook_cloudformation.yml

Created September 7, 2019 18:41

Medium pub

	aws cloudformation deploy /
	--template-file cloudformation_notebook.yaml /
	--stack-name notebook-tutorial

vanbrands / cloudformation_notebook.yaml

Last active September 7, 2019 18:40

pub_cloudformation_notebook

	AWSTemplateFormatVersion: '2010-09-09'
	Description: 'Cloudformation simples para subir um notebook no Sagemaker'
	Resources:
	NotebookTutorial:
	Type: AWS::SageMaker::NotebookInstance
	Properties:
	NotebookInstanceName: "Notebook Tutorial"
	InstanceType: ml.t2.medium
	RoleArn: String
	VolumeSizeInGB: 20

vanbrands / table_rows.sql

Created May 15, 2019 14:32

[Redshift Management] #admin

	select tab.table_schema,
	tab.table_name,
	tinf.tbl_rows as rows
	from svv_tables tab
	join svv_table_info tinf
	on tab.table_schema = tinf.schema
	and tab.table_name = tinf.table
	where tab.table_type = 'BASE TABLE'
	and tab.table_schema not in('pg_catalog','information_schema')
	and tinf.tbl_rows > 1