ifwtrends: ifwtrends: A package for evaluating data from Google Trends

In johannes97s/ifwtrends: Evaluating Google Trends data for economic forecasts

ifwtrends: A package for evaluating data from Google Trends

Description

The package ifwtrends is based on trendecon and gtrendsR and takes data from Google Trends to support economic forecasting.

Functions

• daily_series(): Creates a daily series for time frames where Google usually doesn't provide daily data. Uses the Chow-Lin-method.

• est_trend(): Estimates a common trend between all Google Trends categories and uses that to trend adjust in some other functions. This is updated on a monthly basis in the package. However, the user may use this function for himself on his own.

• factorR2(): Estimates the R squared between keywords.

• forecast_m(): Makes a monthly forecast on a time series based on the data of some Google Trends data. Does currently not work correctly.

• forecast_q(): Makes a quarterly orecast on a time series based on the data of some Google Trends data. Does currently not work correctly.

• gtpreparation(): Makes a search query and applies a trend adjustment (with the common trend), a seasonal adjustment and, if wished, adds lag variables to the resulting data frame.

• gtsearch(): Makes a simple Google search for either a keyword or a category.

• gtseas_adj(): Seasonal adjusts a given time series or uses gtsearch() internally to seasonal adjust a search query. The seasonal adjustment is done via X-13-ARIMA-SEATS from seasonal::seas().

• gttrend_adj(): Trend adjusts a given time series or uses gtsearch() internally to trend adjust a search query. It either can use first differences, a moving average or the common trend computed by est_trend() to adjust the time series.

• pca(): Does a Principal Component Analysis on a search query.

• roll(): Creates some rolling Google search queries for different time frames.

• simple_daily_series(): A more simplistic alternativ for daily_series() that doesn't rely on inner functions from other packages.

For in-depth information and showcases, see the German vignette at vignette(topic = "ifwtrends-demo", package = "ifwtrends")

johannes97s/ifwtrends documentation built on Oct. 9, 2022, 7:01 p.m.