brinnaebent

from distutils.core import setup

setup(name='packagename',
      packages = ['packagename'],
      version='0.1',
      description='Brief Description HERE',
      url='https://github.com/yourusername/packagename',
      download_url = 'https://github.com/yourusername/packagename/archive/0.1.tar.gz', #FILL IN LATER
      author='YOU',
      author_email='YOUR EMAIL',

brinnaebent

import numpy as np
import pandas as pd

'''
Here is a basic description of what these functions do!
Find more information at this link: Link to your GitHub/webpage here
        
'''

brinnaebent

    """
        Brief description of what this function does. 
        Args:
          argument1 (argument type, ie string): description of argument
          argument2 (argument type, ie string): description of argument
        Returns:
            return1 (argument type, ie string): description of return
    """

brinnaebent

$cd directory
$git clone https://github.com/yourname/nameofpackage.git

brinnaebent

$ pip install twine ## Only the first time you do this

#Repeat this every time you update your directory
$ cd local_package_directory
$ python setup.py sdist
$ twine upload dist/*

brinnaebent

[metadata]
description-file = README.md

brinnaebent

$ cd directory
$ git clone https://github.com/username/username.github.io.git

brinnaebent

import numpy as np
import pandas as pd

def LOOCV_featureselection(data, ids, outcomevar, dropcols, idcolumn, numestimators=1000):

brinnaebent

# Separate data for leave-one-person-out-cross-validation (LOOCV)
LOOCV_O = ids 
data[idcolumn] = data[idcolumn].apply(str)
data_filtered = data[data[idcolumn] != LOOCV_O]
data_cv = data[data[idcolumn] == LOOCV_O]
     

brinnaebent

# Train data - all other people in dataframe
data_train = data_filtered.drop(columns=dropcols)
X_train = data_train.drop(columns=[outcomevar])
feature_list = list(X_train.columns) #Get a feature list to use for importances later on

X_train= np.array(X_train) #Re-format as numpy array for input into model
y_train = np.array(data_train[outcomevar]) #Outcome variable here