I'm currently trying to implement a MLR in Python and am not sure how I go about applying the coefficients I've found to future values.
import pandas as pd
import statsmodels.formula.api as sm
import statsmodels.api as sm2
TV = [230.1, 44.5, 17.2, 151.5, 180.8]
Radio = [37.8,39.3,45.9,41.3,10.8]
Newspaper = [69.2,45.1,69.3,58.5,58.4]
Sales = [22.1, 10.4, 9.3, 18.5,12.9]
df = pd.DataFrame({'TV': TV,
'Radio': Radio,
'Newspaper': Newspaper,
'Sales': Sales})
Y = df.Sales
X = df[['TV','Radio','Newspaper']]
X = sm2.add_constant(X)
model = sm.OLS(Y, X).fit()
>>> model.params
const -0.141990
TV 0.070544
Radio 0.239617
Newspaper -0.040178
dtype: float64
So let's say I want to predict out "sales" for the following DataFrame:
EDIT
TV Radio Newspaper Sales
230.1 37,8 69.2 22.4
44.5 39.3 45.1 10.1
... ... ... ...
25 15 15
30 20 22
35 22 36
I've been trying a method I found here but I can't seem to get it working: Forecasting using Pandas OLS
Thank you!
解决方案
Assuming df2 is your new out of sample DataFrame:
model = sm.OLS(Y, X).fit()
new_x = df2.loc[df.Sales.notnull(), ['TV', 'Radio', 'Newspaper']].values
new_x = sm2.add_constant(new_x) # sm2 = statsmodels.api
y_predict = model.predict(new_x)
>>> y_predict
array([ 4.61319034, 5.88274588, 6.15220225])
You can assign the results directly to df2 as follows:
df2.loc[:, 'Sales'] = model.predict(new_x)
To fill missing Sales values from the original DataFrame with predictions from your regression, try:
X = df.loc[df.Sales.notnull(), ['TV', 'Radio', 'Newspaper']]
X = sm2.add_constant(X)
Y = df[df.Sales.notnull()].Sales
model = sm.OLS(Y, X).fit()
new_x = df.loc[df.Sales.isnull(), ['TV', 'Radio', 'Newspaper']]
new_x = sm2.add_constant(new_x) # sm2 = statsmodels.api
df.loc[df.Sales.isnull(), 'Sales'] = model.predict(new_x)