Python random forest ML model wth tkinter gui - python-3.x

Question

I am very new to programming would really appreciated some help. I have developed a ML model using RandomForest. My code works fine when I run my model standalone, but when I try to send data to my code using Tkinter wrapper function I get the value error.
My model code is as follows
import pandas as pd
import numpy as np
from sklearn.model_selection import train_test_split
from sklearn.model_selection import StratifiedShuffleSplit
from sklearn.pipeline import Pipeline
from sklearn.preprocessing import StandardScaler
from sklearn.impute import SimpleImputer
from joblib import dump, load
from sklearn.ensemble import RandomForestRegressor
housing = pd.read_csv("housing_data.csv")
train_set, test_set = train_test_split(housing, test_size=0.2, random_state=42)
splitter = StratifiedShuffleSplit(n_splits = 1, test_size = 0.2, random_state=42)
for train_index, test_index in splitter.split(housing, housing['CHAS']):
strat_train_set = housing.loc[train_index]
strat_test_set = housing.loc[test_index]
median = housing["RM"].median()
housing["RM"].fillna(median)
housing = strat_train_set.drop("MEDV", axis = 1)
housing_labels = strat_train_set["MEDV"].copy()
my_pipeline = Pipeline([
('imputer', SimpleImputer(strategy = "median")), #imputer takes the median value and fills all the blank fields
('std_scaler', StandardScaler()),
])
housing_pip = my_pipeline.fit_transform(housing)
print(housing_pip.shape)
model_random_forest = RandomForestRegressor()
model_random_forest.fit(housing_pip, housing_labels)
dump(model_random_forest, 'Dragon.joblib')
I am using a fucntion to enter the values and get the predictions. Function code is as follows
a=[]
b=[]
c=[]
d=[]
e=[]
f=[]
g=[]
h=[]
i=[]
j=[]
k=[]
l=[]
m=[]
def predictor(array = np.array([[a],[b],[c],[d],[e],[f],[g],[h],[i],[j],[k],[l],[m]])):
print(model_random_forest.predict(array))
predictor([[1,2,3,4,5,6,7,8,9,10,11,12,13]])
Code works absolutely fine but when I use Tkinter GUI wrapper function to send the values to this function it gives ValueError: Number of features of the model must match the input. Model n_features is 13 and input n_features is 1
My Tkinter wrapper function is as follows
def make_predict():
test.predictor([[crim.get()],[zn.get()],[indus.get()],[chas.get()],[nox.get()],[rm.get()],[age.get()],[dis.get()],[rad.get()],[tax.get()],[ptratio.get()],[b.get()],[lstat.get()]])
crim.get, zn.get etc. are entry boxes in tkinter. Guys I would really appreciate your help. I am new to programming and have already spen lots of time solving this error.
Just in case I am also pasting my complete Tkinter code below
from tkinter import*
import numpy as np
import test
window = Tk()
def make_predict():
test.predictor([[crim.get()],[zn.get()],[indus.get()],[chas.get()],[nox.get()],[rm.get()],[age.get()],
[dis.get()],[rad.get()],[tax.get()],[ptratio.get()],[b.get()],[lstat.get()]])
#t1.delete(0,END)
#t1.insert(END,(model_random_forest.predict(features)))
window.title("Property Price Predictor")
window.geometry('500x800')
mainLabel = Label(window, text="Property Price Predictor", font=("Arial Bold", 20))
mainLabel.grid(row=2, columnspan=8)
crim=StringVar()
e1 = Entry(window, width = 25, textvariable=crim) #crim
e1.grid(row = 3, column = 2)
l1 = Label(window, text = 'PER CAPITA CRIME RATE BY TOWN')
l1.grid(row = 4, column = 2)
zn = StringVar()
e2 = Entry(window, width = 25, textvariable=zn) #ZN
e2.grid(row = 6, column = 2)
l2 = Label(window, text = 'PROPORTION OF RESIDENTIAL LAND ZONED FOR LOTS OVER 25,000 SQ.FT.')
l2.grid(row = 7, column = 2)
indus=StringVar()
e3 = Entry(window, width = 25, textvariable=indus) #INDUS
e3.grid(row = 9, column = 2)
l3 = Label(window, text = 'PROPORTION OF NON-RETAIL BUSINESS ACRES PER TOWN')
l3.grid(row = 10, column = 2)
chas=StringVar()
e4 = Entry(window, width = 25, textvariable=chas) #CHAS
e4.grid(row = 12, column = 2)
l4 = Label(window, text = 'CHARLES RIVER DUMMY VARIABLE (= 1 IF TRACT BOUNDS RIVER; 0 OTHERWISE)')
l4.grid(row = 13, column = 2)
nox=StringVar()
e5 = Entry(window, width = 25, textvariable=nox) #NOX
e5.grid(row = 15, column = 2)
l5 = Label(window, text = 'NITRIC OXIDES CONCENTRATION (PARTS PER 10 MILLION)')
l5.grid(row = 16, column = 2)
rm=StringVar()
e6 = Entry(window, width = 25, textvariable=rm) #RM
e6.grid(row = 18, column = 2)
l6 = Label(window, text = 'AVERAGE NUMBER OF ROOMS PER DWELLING')
l6.grid(row = 19, column = 2)
age=StringVar()
e7 = Entry(window, width = 25, textvariable=age) # AGE
e7.grid(row = 21, column = 2)
l7 = Label(window, text = 'PROPORTION OF OWNER-OCCUPIED UNITS BUILT PRIOR TO 1940')
l7.grid(row = 22, column = 2)
dis=StringVar()
e8 = Entry(window, width = 25, textvariable=dis) #DIS
e8.grid(row = 24, column = 2)
l8 = Label(window, text = 'WEIGHTED DISTANCES TO FIVE BOSTON EMPLOYMENT CENTRES')
l8.grid(row = 25, column = 2)
rad=StringVar()
e9 = Entry(window, width = 25, textvariable=rad) #RAD
e9.grid(row = 27, column = 2)
l9 = Label(window, text = 'INDEX OF ACCESSIBILITY TO RADIAL HIGHWAYS')
l9.grid(row = 28, column = 2)
tax=StringVar()
e10 = Entry(window, width = 25, textvariable=tax ) #TAX
e10.grid(row = 30, column = 2)
l10 = Label(window, text = 'FULL-VALUE PROPERTY-TAX RATE PER $10,000')
l10.grid(row = 31, column = 2)
ptratio=StringVar()
e11 = Entry(window, width = 25, textvariable=ptratio) #PTRATIO
e11.grid(row = 33, column = 2)
l11 = Label(window, text = 'PUPIL-TEACHER RATIO BY TOWN')
l11.grid(row = 34, column = 2)
b=StringVar()
e12 = Entry(window, width = 25, textvariable=b) #B
e12.grid(row = 36, column = 2)
l12 = Label(window, text = '1000(BK - 0.63)^2 WHERE BK IS THE PROPORTION OF BLACKS BY TOWN')
l12.grid(row = 37, column = 2)
lstat=StringVar()
e13 = Entry(window, width = 25, textvariable=lstat) #LSTAT
e13.grid(row = 39, column = 2)
l13 = Label(window, text = '% LOWER STATUS OF THE POPULATION')
l13.grid(row = 40, column = 2)
t1 = Text(window, height = 1, width = 20)
t1.grid(row = 45, column = 2)
l15 = Label(window, text = 'PREDICTION')
l15.grid(row = 46, column = 2)
db1 = Button(window, text = 'PREDICT', width=20, command = make_predict)
db1.grid(row=48, column=2)
window.mainloop()

Answer 1

Your housing dataset most probably has missing values
you need to drop the columns with missing values instead of imputing them.
Here is the code:
First check for the missing values in your dataset
housing_train = X_train
missing_values_columns = [col for col in housing_train.columns
if housing_train[col].isnull().any()]
housing_transformed = X_test.drop(missing_values_columns, axis=1
model_random_forest = RandomForestRegressor()
model_random_forest.fit(housing_transformed, housing_labels)