I need to generate an animating Heatmap but not working?

Using Streamlit
2

You are in the right direction generating the frames that will compose the animation. However, the trick of sequentially throwing single pictures to the screen won’t do the trick (:melting_face: Matlab flashbacks), you’ll need to throw them to a video encoder. Here is an example using the matplotlib.animation module:

matplotlibanimate

Code:

import matplotlib.pyplot as plt
from matplotlib.animation import FuncAnimation
import numpy as np
import streamlit as st
from itertools import cycle

## Generate some data
randomdata = np.random.random(size=(5,3,3))
harvest_cycle = cycle([i for i in randomdata])
harvest = next(harvest_cycle)

## Initialize figure
fig,ax = plt.subplots(figsize=[4,4])
ax.set_aspect('equal')
heatmap = ax.imshow(harvest)

st.markdown("## :camera: Initial condition")
st.pyplot(fig)

## Animate 
def update(frame_number): 
    ax.cla() ## Clears the data passed to the axes
    ##  This is important so the figure does not store data
    ##  of every single frame and slows everything down
    
    ax.imshow(next(harvest_cycle))
    ax.set_title(f"Frame: {frame_number}",fontsize=24)

animation = FuncAnimation(fig,update,frames=50)

## Saved mp4 video
animation.save("movie.mp4")

## Keep jshmtl representation
animjs = animation.to_jshtml()

st.markdown("## :movie_camera: Enconded MP4 movie")
st.video("movie.mp4")

st.markdown("## :control_knobs: Using `jshtml` from matplotlib")
st.components.v1.html(animjs,height=600)
3 Likes