iris.py

# S10.1: Copy this code cell in 'iris_app.py' using the Sublime text editor. You have already created this ML model in the previous class(es).
# Importing the necessary libraries.
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import streamlit as st
from sklearn.model_selection import train_test_split
from sklearn.svm import SVC
from sklearn.linear_model import LogisticRegression
from sklearn.ensemble import RandomForestClassifier

# Loading the dataset.
iris_df = pd.read_csv("iris-species.csv")

# Adding a column in the Iris DataFrame to resemble the non-numeric 'Species' column as numeric using 'map()' function.
# Creating the numeric target column 'Label' to 'iris_df' using 'map()'.
iris_df['Label'] = iris_df['Species'].map({'Iris-setosa': 0, 'Iris-virginica': 1, 'Iris-versicolor':2})

# Creating a model for Support Vector classification to classify the flower types into labels '0', '1', and '2'.

# Creating features and target DataFrames.
X = iris_df[['SepalLengthCm','SepalWidthCm', 'PetalLengthCm', 'PetalWidthCm']]
y = iris_df['Label']

# Spliting the data into training and testing sets.
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.33, random_state = 42)

# Creating the SVC model and storing the accuracy score in a variable 'score'.
svc_model = SVC(kernel = 'linear')
svc_model.fit(X_train, y_train)
score = svc_model.score(X_train, y_train)


# Creating a Logistic Regression model.
rf_clf = RandomForestClassifier(n_jobs = -1, n_estimators = 100)
rf_clf.fit(X_train, y_train)

# Creating a Random Forest Classifier model.
log_reg = LogisticRegression(n_jobs = -1)
log_reg.fit(X_train, y_train)

@st.cache()
def prediction(model, sepal_length, sepal_width, petal_length, petal_width):
  species = model.predict([[sepal_length, sepal_width, petal_length, petal_width]])
  species = species[0]
  if species == 0:
    return "Iris-setosa"
  elif species == 1:
    return "Iris-virginica"
  else:
    return "Iris-versicolor"
  

# S2.3: Copy the following code and paste it in the 'improved_iris_app.py' file after the previous code.
# Add title widget
st.sidebar.title("Iris Species Prediction App")

# Add 4 sliders and store the value returned by them in 4 separate variables.
s_len = st.sidebar.slider("Sepal Length", float(iris_df["SepalLengthCm"].min()), float(iris_df["SepalLengthCm"].max()))
# The 'float()' function converts the 'numpy.float' values to Python float values.
s_wid = st.sidebar.slider("Sepal Width", float(iris_df["SepalWidthCm"].min()), float(iris_df["SepalWidthCm"].max()))
p_len = st.sidebar.slider("Petal Length", float(iris_df["PetalLengthCm"].min()), float(iris_df["PetalLengthCm"].max()))
p_wid = st.sidebar.slider("Petal Width", float(iris_df["PetalWidthCm"].min()), float(iris_df["PetalWidthCm"].max()))

# Add a select box in the sidebar with the 'Classifier' label.
# Also pass 3 options as a tuple ('Support Vector Machine', 'Logistic Regression', 'Random Forest Classifier').
# Store the current value of this slider in the 'classifier' variable.
classifier = st.sidebar.selectbox('Classifier', ('Support Vector Machine', 'Logistic Regression', 'Random Forest Classifier'))

# When the 'Predict' button is clicked, check which classifier is chosen and call the 'prediction()' function.
# Store the predicted value in the 'species_type' variable accuracy score of the model in the 'score' variable.
# Print the values of 'species_type' and 'score' variables using the 'st.text()' function.
if st.sidebar.button("Predict"):
  if classifier == 'Support Vector Machine':
    species_type = prediction(svc_model, s_len, s_wid, p_len, p_wid)
    score = svc_model.score(X_train, y_train)

  elif classifier =='Logistic Regression':
    species_type = prediction(log_reg, s_len, s_wid, p_len, p_wid)
    score = log_reg.score(X_train, y_train)

  else:
    species_type = prediction(rf_clf, s_len, s_wid, p_len, p_wid)
    score = rf_clf.score(X_train, y_train)

  st.write("Species predicted:", species_type)
  st.write("Accuracy score of this model is:", score)