- R (3.6.1)
- RStudio (1.2.5019)
- Required packages
To install required R packages, run script setup.R
.
data
: Raw and processed data.plots
: Output plots of correlation matrices, clustering dendrogram, linear models, machine learning models.outputs
: CSV/Text outputs.settings
: Common setting of the project.scripts
: Main script for analysis and data modeling.utils
: Common ultility functions.
All global settings are placed in settings/settings.R
.
The default time-series anlysis is using data of 30 seconds before to predict the class of next 5 seconds. You can change these values for different analysis.
TIME_PREV_SECONDS <- 30 # [5, 10, 15, 30]
TIME_NEXT_SECONDS <- 5
As the project is using Plotly as a visualization tool, you can set up your Plotly account in settings/settings.R
.
PLOTLY_USERNAME <- "<Your Plotly Username>"
PLOTLY_API_KEY <- "<Your Plotly API Key>"
The Plotly lib needs orca
as an static image exporting backend.
Follow instruction at https://github.com/plotly/orca for installation.
- Run
scripts/correlation.R
- Run
scripts/clustering.R
- Run
scripts/linearModel.R
http://jef.works/R-style-guide/
In RStudio, run following commands:
install.packages("styler")