Sampling distributions Shiny app - Shiny Contest Submission

Sampling distributions Shiny app

Authors: Tomas Capretto

Abstract: Shiny app to visualize sampling distributions of several statistics computed with samples obtained from arbitrary mixtures of distributions.

Full Description:

Basic usage

The objective of this application is to enable people to explore sampling
distributions of several statistics computed with samples coming from arbitrary
mixtures of probability distributions.

Add distributions to the mixture.

Choose one distribution at a time from the input on the left panel and click on +.
This will add the distribution to the mixture and insert inputs to tune its
parameters as well as the weight it has in the mixture. You can repeat this process
as many times as you want, adding a new distribution to the mixture in each time.

Tune sampling parameters

  • Sample size is the number of draws we take from the (theoretical) mixture
    we construct. You can specify any integer number between 2 and 1000.
  • Repetitions indicates how many times we repeat the sampling process. The larger
    this number the better the approximation of the empirical sampling distribution.
    This must be between 10 and 1000.
  • Statistic determines what statistic we compute with the sampled values.
    Currently, these are the available options:
    • Mean
    • Median
    • Minimum
    • Maximum
    • Percentile
  • Percentile indicates which percentile we compute when we select Percentile option
    in Statistic.

Explore the charts

Just see the interactive charts and make up your conclusions!


Keywords: statistics, sampling-distributions, shiny.semantic, echarts4r
Shiny app: https://tomicapretto.shinyapps.io/samplingdistributions
Repo: GitHub - tomicapretto/samplingdistributions: Sampling distributions shiny app
RStudio Cloud: RStudio Cloud

Thumbnail:
image

Full image: