Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Introduce default number of trials to visualization features.

See original GitHub issue

Motivation

Visualization features are unstable with more than hundreds/thousands of trials. For example, the following program stoped working during visualization in my environment.

def objective(trial):
    x = trial.suggest_uniform('x', 0, 1)
    y = trial.suggest_uniform('y', 0, 1)

    return x + y
    
study = optuna.create_study()
study.optimize(objective, n_trials=5000)
optuna.visualization.plot_contour(study)

Description

Just plotting the latest trials would mitigate this problem. For example, we can add an optional n_trials argument to the plot functions so that they plot the latest 100 trials by default.

def plot_intermediate_values(study, n_trials=100):
    ...  # Plot `n_trials` number of trials instead of all of them.

Issue Analytics

State:
Created 4 years ago
Reactions:1
Comments:7 (2 by maintainers)

Top GitHub Comments

3reactions

michiboocommented, Aug 4, 2020

@hvy Hi sorry for the late response, I will try fix this conflict for this issue today.

3reactions

grfreitascommented, Mar 9, 2020

I want to give a suggestion here: the figures are generated by plotly with go.Scatter, if plotting the graph is the reason for the program stopping, one way to reduce the lag generated by it is by changing the call to go.Scattergl. It is capable of rendering more points with less lag.

https://plot.ly/python/webgl-vs-svg/

Top Results From Across the Web

Chapter 1. The Seven Stages of Visualizing Data - O'Reilly

Whenever we analyze data, our goal is to highlight its features in order of their importance, reveal patterns, and simultaneously show features that...

13 Powerful Methods for Data Visualization Success l Sisense

Eliminate bad graphs and charts! Learn how to find the best way to visualize data and get your message across for your business...

What Is Data Visualization? Definition & Examples | Tableau

Visualize patterns and relationships. Disadvantages. While there are many advantages, some of the disadvantages may seem less obvious. For example, when viewing ...

Domo Data Visualization Simplified: 5 Easy Steps

This article will provide you with an understanding of how you can use Domo Data Visualizations to gather useful insights from business data....

Hands-On Tutorial: Visualization Enhancements

Throughout this tutorial, we will use the default sample (which ... you will explore another visualization enhancement feature: number formatting.