forecast() in R repeats predictions - r

Question

I'm trying to build a forecast to predict future values of a keyword from Google Trends data.
My data is the daily indexed search volume from Jan 1 to Jun 30, 2020 for a keyword, saved in a CSV file:
Date | Keyword
2020-01-01 | 55
2020-01-02 | 79
2020-01-03 | 29
...
2020-06-29 | 19
2020-06-30 | 32
My R code seems to work okay until it generates the forecasts.
library(forecast)
data <- read.csv("<file path>.csv", header=TRUE)
#build time series data
#start and end periods of observed data
inds <- seq(as.Date("2020-01-01"), as.Date("2020-06-30"), by = "day")
#the frequency = 7 days (i.e. week)
sts <- ts(data$Keyword, start = c(2020, as.numeric(format(inds[1], "%j"))), frequency = 7)
#generate the forecast
model.ets <- ets(sts, model = "ANA")
fc.ets <- forecast(model.ets, h = 60)
plot(fc.ets)
The problem I'm having is that the forecast simply repeats the same pattern (doesn't seem to take into account the error, trend and/or seasonality to adjust the predictions).
I think I need to adjust the forecast() function but not sure how to do it.

Answer 1

In this case we have a daily series spanning less than a year that appears to display a weekly seasonality. Please note as is given here: https://otexts.com/fpp2/ts-objects.html [2.1 - ts objects], the frequency given to the ts object is 52.18 which is 365.25/7, the number of weeks in a year (taking into account leap years). This seasonality rules out the use of ets models which can't handle data with frequency greater than 24, unless used in combination with STL (Seasonal and Trend decomposition using Loess). As such I would recommend exploring other models. STL + ETS(A, Ad, N) [2nd best model] point forecasts look most realistic but the range in our prediction intervals is much larger when compared against the TBATS(1, {0,0}, 0.92, {<52.18, 6>}) model [best model] please see and play around with the below:
ts_ausAirBnb <- ts(ausAirBnb$airbnb_australia_, start = min(ausAirBnb$day), frequency = 52.18)
plot(decompose(ts_ausAirBnb))
snaivefit <- snaive(ts_ausAirBnb)
snaivefcast <- forecast(snaivefit, h = 60)
aafit <- auto.arima(ts_ausAirBnb)
aafcast <- forecast(aafit, h = 60)
stlffit <- stlf(ts_ausAirBnb, h = 60)
stlfcast <- forecast(stlffit, h = 60)
stlmfit <- stlm(ts_ausAirBnb)
stlmfcast <- forecast(stlmfit, h = 60)
tbatsfit <- tbats(ts_ausAirBnb)
tbatsfcast <- forecast(tbatsfit, h = 60)
nnetfit <- nnetar(ts_ausAirBnb)
nnetfcast <- forecast(nnetfit, h = 60)
autoplot(snaivefcast)
autoplot(aafcast)
autoplot(etsfcast)
autoplot(stlfcast)
autoplot(stlffit)
autoplot(stlmfcast)
autoplot(tbatsfcast)
autoplot(nnetfcast)