import streamlit as st
from sklearn import datasets
import numpy as np
import pandas as pd

st.title('Iris Flower Prediction App')
st.sidebar.title('Input Parameters')

@st.cache
def get_data():
    print('Load data ...')
    iris = datasets.load_iris()
    return iris.data[:, :2], iris.target

@st.cache(allow_output_mutation=True)
def train_model(train_X, train_Y):
    print('Train model ...')
    from sklearn.ensemble import RandomForestClassifier
    clf = RandomForestClassifier()
    clf.fit(train_X, train_Y)

    return clf

def get_features():
    sepal_length = st.sidebar.slider('Sepal length', 4.3, 7.9, 5.4)
    sepal_width = st.sidebar.slider('Sepal width', 2.0, 4.4, 3.4)
    data = {'sepal_length': sepal_length,
            'sepal_width': sepal_width}
    features = pd.DataFrame(data, index=[0])
    return features

X, Y = get_data()
clf = train_model(X, Y)

df = get_features()
prediction = clf.predict(df)

def display_image(value):
    from PIL import Image

    if value == 0:
        image = Image.open('Iris-setosa.jpg')
        st.image(image, caption='Iris Setosa', width=200)
    elif value == 1:
        image = Image.open('IRIS_VERSICOLOR.jpeg')
        st.image(image, caption='Iris Versicolor', width=200)
    else:
        image = Image.open('Iris_virginica.jpg')
        st.image(image, caption='Iris Virginica', width=200)

display_image(prediction[0])
		

To avoid some code is re-run

@st.cache(persist=True)
def get_data():
    print('Load data '+str(count)+' ...')
    iris = datasets.load_iris()
    return iris.data, iris.target
		

Widget

• button
• checkbox
• radio
• sidebar
• selectbox
• multiselect
• slider
• text_input
• number_input
• text_area
• date_input
• time_input
• file_uploader
• beta_color_picker

Writing

• title, set the app’s title
• header, heading level
• subheader, sub heading level
• text, write pure text
• markdown, write markdown text
• latex, display mathematical expressions formatted as LaTeX
• write, write different things depending on what the input is
• dataframe, display a dataframe as an interactive table
• table, display a static table
• json, display object or string as a pretty-printed JSON string

Plotting

• line_chart
• area_chart
• bar_chart
• pyplot
• altair_chart
• vega_lite_chart
• plotly_chart, display an interactive Plotly chart
• bokeh_chart, display an interactive Bokeh chart
• pydeck_chart, supports 3D maps, point clouds, and more
• deck_gl_chart, follows Deck.GL’s JavaScript API
• graphviz_chart, display a graph using the dagre-d3 library
• map, display a map with points on it
• image, display an image or list of images
• audio, display an audio player
• video, display a video player

Sidebar

• st.sidebar.[element_name]

streamlit run [appname]
		

Local Project

1. Create a project
2. pipreq projectDirectory, create requirements.txt
3. Create setup.sh

mkdir -p ~/.streamlit/

echo "\
[general]\n\
email = \"your-email@domain.com\"\n\
" > ~/.streamlit/credentials.toml

echo "\
[server]\n\
headless = true\n\
enableCORS=false\n\
port = $PORT\n\
" > ~/.streamlit/config.toml
		

4. Create Procfile, used to first execute the setup.sh and then call streamlit run to run the application

web: sh setup.sh && streamlit run app.py
		

Deploy at Heroku

1. Create a app at Heroku
2. git init
3. heroku git:remote -a [appname]
4. git add .
5. git commit -am "comments"
6. git push heroku master
7. App at Heroku

Deploying your Streamlit dashboard with Heroku

API