Forecast Simple Models

Overview

This shiny application is for testing some simple forecasting models. We used the Google Trends database for this application using the “gtrendsR” package. Instructions

Shiny Application

Github

Instructions

To use this application, enter a keyword that interests you, select the forecast template and click on “Run Code”.

Results

Plot Output: See the hits for the keyword searched in Google Trends and the forecasting for this search. The black curve within the predictive area is the training curve.

Summary: Indicates the calculated errors for the applied forecast. This app will have new updates inserting new templates and views.

1) Using the gtrendsR package, the user can query some word in Google Trends database.

2) After that he can choose a Forecast Model (Forecast Package) and click in a Run Code Button.

Example:

library(gtrendsR)
library(forecast)
library(tseries)
search <- gtrends("Rstudio")
gt.fc <- search$interest_over_time
fc <- meanf(gt.fc$hits)