Let's break down your problem and answer your questions step by step. --- ## **The Error Explained** Your error is: ``` Error in plot.default(...) : formal argument "type" matched by multiple actual arguments ``` This happens because you are calling: ```r plot(BikeDemand, type = "l", col = 1, ylim = c(0, 35), main = "Types of moving averages", sub = "Oct 3 2016--Oct 23 2016", xlab = "Hours", ylab = "Bike Usage/Demand") ``` **Explanation:** `BikeDemand` is a data frame with 7 columns (time + 6 stations). When you call `plot(BikeDemand, ...)`, the `plot()` function dispatches to `plot.data.frame()`. This function does *not* accept the `type` argument (which is for plotting lines in `plot.default`), leading to a conflict. --- ### **How to Fix** #### **Option 1: Plot Each Station Individually** If you want to plot time series of the 6 stations, you should plot each column (each station) **against time**. ```r # Suppose time is in the first column, stations in columns 2-7 matplot(BikeDemand[[1]], BikeDemand[,-1], type = "l", ylim = c(0, 35), col = 1:6, lty = 1, main = "Types of moving averages", sub = "Oct 3 2016--Oct 23 2016", xlab = "Hours", ylab = "Bike Usage/Demand") legend("topright", legend = colnames(BikeDemand)[-1], col = 1:6, lty = 1) ``` - `matplot()` is designed to plot multiple columns against a single x-axis. - `BikeDemand[[1]]` is your time column. - `BikeDemand[,-1]` are your 6 station columns. #### **Option 2: Plot a Single Station** If you want to plot, for example, station 1: ```r plot(BikeDemand[[1]], BikeDemand[[2]], type = "l", ylim = c(0, 35), main = "Station 1 Bike Demand", xlab = "Hours", ylab = "Bike Usage/Demand") ``` --- ## **Answering Your Questions** ### **1. Predict the Next 3 Hours of Each Station** You can use a simple moving average, ARIMA, or other time series models. For illustration, let's use the built-in `forecast` package: #### **Using ARIMA (auto.arima) for Each Station** ```r library(forecast) # Predict next 3 hours for each station preds <- list() for (i in 2:7) { station_ts <- ts(BikeDemand[[i]], frequency = 24) # hourly, daily seasonality fit <- auto.arima(station_ts) forecast_vals <- forecast(fit, h = 3) print(paste("Station", colnames(BikeDemand)[i], "next 3 hours:")) print(forecast_vals$mean) preds[[colnames(BikeDemand)[i]]] <- forecast_vals } ``` **Discussion:** After running this code, look at the printed `forecast_vals$mean` for each station. These are your predictions for the next 3 hours. --- ### **2. Aggregating Hourly Data into Daily Data, Predict Next 3 Days** Aggregate to daily: ```r # Remove time column, aggregate by day (24 hours per day) daily_data <- aggregate(BikeDemand[,-1], by = list(rep(1:(nrow(BikeDemand)/24), each=24)), FUN = sum) # Remove grouping column daily_data <- daily_data[,-1] # Now, daily_data has 20 rows (days) x 6 columns (stations) # Predict next 3 days for each station for (i in 1:6) { station_ts <- ts(daily_data[[i]], frequency = 7) # daily, weekly seasonality fit <- auto.arima(station_ts) forecast_vals <- forecast(fit, h = 3) print(paste("Station", colnames(daily_data)[i], "next 3 days:")) print(forecast_vals$mean) } ``` **Discussion:** The printed `forecast_vals$mean` for each station gives your 3-day forecasts. --- ## **Summary** - Your plotting error is due to using `plot()` on a data frame with the `type` argument; use `matplot()` for multiple series. - For forecasting, use `auto.arima()` from the `forecast` package for a quick approach, both for hourly and aggregated daily data. - Interpret the forecast output to discuss expected demands. --- **If you need more information on interpreting the forecasts or want to use your custom moving average function for prediction, let me know!**