app.py

import pickle

import matplotlib.pyplot as plt
import pandas as pd
import seaborn as sns
import streamlit as st
from sklearn.ensemble import RandomForestClassifier
from sklearn.metrics import accuracy_score
from sklearn.model_selection import train_test_split

st.title("Penguin Classifier")

st.write(
    """App uses 6 inputs to predict
     the species of penguin using a model
     built on the Palmer's Penguins dataset.
    """
)

# password 설정
password_guess = st.text_input("Input Password?")
if password_guess != "streamlit":
    st.stop()

# updata csvfile
penguin_file = st.file_uploader("Upload your own data")

if penguin_file is None:
    rf_pickle = open("models/rf_penguin.pickle", "rb")
    map_pickle = open("models/class_penguin.pickle", "rb")
    rfc = pickle.load(rf_pickle)
    unique_penguin_mapping = pickle.load(map_pickle)
    rf_pickle.close()
    map_pickle.close()
    penguin_df = pd.read_csv("penguins.csv")
else:
    penguin_df = pd.read_csv(penguin_file)
    penguin_df['sex'].fillna(penguin_df['sex'].mode()[0], inplace=True)
    penguin_df = penguin_df.dropna()
    output = penguin_df["species"]
    features = penguin_df[
        [
            "island",
            "bill_length_mm",
            "bill_depth_mm",
            "flipper_length_mm",
            "body_mass_g",
            "sex",
        ]
    ]
    features = pd.get_dummies(features)
    output, unique_penguin_mapping = pd.factorize(output)
    x_train, x_test, y_train, y_test = train_test_split(features, output, test_size=0.8)
    rfc = RandomForestClassifier(random_state=15)
    rfc.fit(x_train.values, y_train)
    y_pred = rfc.predict(x_test.values)
    score = round(accuracy_score(y_pred, y_test), 2)
    
    rf_pickle = open("models/rf_penguin.pickle", "wb")
    pickle.dump(rfc, rf_pickle)
    rf_pickle.close()

    output_pickle = open("models/class_penguin.pickle", "wb")
    pickle.dump(unique_penguin_mapping, output_pickle)
    output_pickle.close()
    st.write(
        f"""Trained a Random Forest model on these data,
        it has a score of {score}! """
    )

# selectbox, button 만들기    
with st.form("user_inputs"):
    island = st.selectbox("Penguin Island", options=["Biscoe", "Dream", "Torgerson"])
    sex = st.selectbox("Sex", options=["Female", "Male"])
    bill_length = st.number_input("Bill Length (mm)", min_value=0)
    bill_depth = st.number_input("Bill Depth (mm)", min_value=0)
    flipper_length = st.number_input("Flipper Length (mm)", min_value=0)
    body_mass = st.number_input("Body Mass (g)", min_value=0)
    st.form_submit_button()

island_biscoe, island_dream, island_torgerson = 0, 0, 0
if island == "Biscoe":
    island_biscoe = 1
elif island == "Dream":
    island_dream = 1
elif island == "Torgerson":
    island_torgerson = 1

sex_female, sex_male = 0, 0
if sex == "Female":
    sex_female = 1
elif sex == "Male":
    sex_male = 1

# Predction    
new_prediction = rfc.predict(
    [
        [
            bill_length,
            bill_depth,
            flipper_length,
            body_mass,
            island_biscoe,
            island_dream,
            island_torgerson,
            sex_female,
            sex_male,
        ]
    ]
)
st.subheader("Predicting Your Penguin's Species:")
prediction_species = unique_penguin_mapping[new_prediction][0]
st.write(f"# Prediction Species: **{prediction_species}")
st.write(
    """Machine learning
    (Random Forest) model to predict the
    species, the features used in this
    prediction are ranked by relative
    importance below."""
)

st.write(
    """Below are the histograms for each
continuous variable separated by penguin species.
The vertical line represents the inputted value."""
)

fig, ax = plt.subplots()
ax = sns.displot(x=penguin_df["bill_length_mm"], hue=penguin_df["species"])
plt.axvline(bill_length)
plt.title("Bill Length by Species")
st.pyplot(ax)

fig, ax = plt.subplots()
ax = sns.displot(x=penguin_df["bill_depth_mm"], hue=penguin_df["species"])
plt.axvline(bill_depth)
plt.title("Bill Depth by Species")
st.pyplot(ax)

fig, ax = plt.subplots()
ax = sns.displot(x=penguin_df["flipper_length_mm"], hue=penguin_df["species"])
plt.axvline(flipper_length)
plt.title("Flipper Length by Species")
st.pyplot(ax)