sagivba / machinelearningutils Goto Github PK

class LinearModelUtils
def get_formula
def rmse
def rmsle
def mae
def model_info_ record
return ModelInfoRecord

class ModelInfo(list)

class ModelInfoRecord
def init

• self
• model_name
• col_list
• actual_lbl
• formula
  -prediction_column
• clf
  class LinearModelInfoRecord(ModelInfoRecord)
  def init
• self
  -rmse
  -rmsle
  -mae
  -formula
  class ModelInfoReport

linear regression:
better_dummies:
splitting values by ranges or function
for example:
(
column: hour,
{
rush_hours : [7,8,9, ],
back_hour : [15,16,17]
others: [1:6. 10:14, 18:24]
}
)

will split the "hour" column in our df into 3 columns : rush_hours ,back_hour,others

from sklearn import datasets
from MachineLearningUtils.UsefulPlots import VisPlotPlayGround
from matplotlib import cm
from matplotlib import pyplot as plt
import pandas as pd
import numpy as np

def main():

iris = datasets.load_iris()
_df=data1 = pd.DataFrame(data= np.c_[iris['data'], iris['target']],
                 columns= iris['feature_names'] + ['target'])
print (_df.info())
# exit()
plotter=VisPlotPlayGround(df=_df, ggplot=True, cmap=cm.jet)
cmaps = [m for m in cm.datad if not m.endswith("_r")]
print(sorted(cmaps))
for c in sorted(cmaps):
    plotter.show_colormap(c)
plt.show()

if name == 'main':
main()

http://scikit-learn.org/stable/modules/learning_curve.html

def init(self,df,targert_name)
def plot_releshenships()
"""
smart plot the relashenships between target and ther colums
"""

add examples of small plots in the main readme file
add more examples in the example directory with readme file cointaining some short eamples

aget fig and matrix of axess and return fig

Don't test visualization, test different datasets using None and np.NaN
Test for the returned object and their attributes.
Check that the Utils does not throw exceptions

from sklearn.metrics import classification_report
from sklearn.metrics import accuracy_score

methods:
add_columns_one_to_many(self,df_master,df_details,function,master_columns_lst,details_columns_lst)
for row in df_master[master_columns_lst,:]:
df_details[details_columns_lst] where joined with row
add columns to master_columns_lst function(row,)

 returns df_master with_new_columns

def plot_confusion_matrix(self,
cm, classes,
normalize=True,
title='Confusion matrix',
cmap=plt.cm.Blues):

This method should prints and plots the confusion matrix.
Normalization can be applied by setting normalize=True.

• split_to_columns(column,split_func,splited_columns_name) : returns tuple of colomns

• drop_columns(df,columns_list) returns df without the columns_list

• apply_by_dict(df, **kward)
  for key, value in kwargs.iteritems():
  if key not in list(df):
  print warn not in columns list
  continue
  if value is function:
  value=[value]
  if value int iterable:
  print warn
  for f in value
  df[key]=df[key].apply(f)
  return df