shiny_hidden_markov_analysis: Hidden Markov for Shiny

In brentscott93/biophysr: Analysis tools for muscle biophysics

Description

Hidden Markov for Shiny

Usage

shiny_hidden_markov_analysis(
  trap_selected_date,
  mv2nm,
  nm2pn,
  overlay_color,
  file_type,
  hm_emcontrol
)

Arguments

trap_selected_date	absolute file path
mv2nm	numeric conversion
nm2pn	numeric conversion
overlay_color	color choice for model overlay
file_type	either 'csv' or 'txt'
hm_emcontrol	logical. TRUE/FALSE

Value

Creates a folder named 'results' in each 'obs-##' folder and saves files directly to it

Examples

This is an interactive plot of the running variance and running mean.
The algorithm for event detection is a Hidden Markov Model and these are the data the model receives as input. The model is fitted with the EM algorithm and the
gray shaded regions are the binding events identified through state sequence decoding via the Viterbi alogorithm.




Below is an interactive plot of the raw data (model does not use this data).
The overlay is the Hidden Markov Model 'state' prediction multipled by several conversion factors.
These convert the x-axis (time) from 'windows' to 'data points' to 'seconds'.
The HMM state, baseline mean, and measured step size are used to scale the model overlay to each step and subsequent baseline level.


Plots of the running mean vs. running variance.
This provides insight into how the model divided data into either the baseline or event populations.

brentscott93/biophysr documentation built on Sept. 14, 2021, 2:35 a.m.