https://classroom.github.com/a/1TdiZRPk

Due: 2020-04-20 by 11:59pm

Use the dataset bikes, which records the number of bikes rented per day by Capital Bikeshare system, in Washington D.C., USA.

  1. Use the variable cnt and a categorical variable of your choice to make the following.

    1. A bivariate plot with at least two layers (geom_x and geom_y) of your choice which describe the population distribution. Provide at least two complete English sentences describing two interesting facts about your data, referencing some statistical aspect of the data amongst the groups in the context of the data.

    2. A bivariate plot with group means and confidence intervals via the ggplot2 layer stat_summary(). Provide at least one complete English sentence describing an interesting fact about your data, referencing some statistical aspect of the data amongst the groups in the context of the data.

    3. Use the libraries boot and dplyr to provide at least two confidence intervals, neither of 95%, interpretted in context of the data with the standard phrase used for confidence intervals.

  2. Repeat 1., a. through c., using the variable cnt and a new categorical variable different from above.

  3. Challenge question – for those who want one.

    Suppose you have code analogous to

    b <- boot(...)

    Fill in the next two arguments to the function apply() to reproduce the confidence intervals calculated by boot.ci(b, type="perc", ...)

    apply(b$t, ...) # fill in the next two arguments

    Note. boot.ci() does some other fancy things such that your answers might be different by a small amount.

    Hint. It might be helpful to first get an understanding for what apply() does and then practice calculating a confidence interval from just one column of b$t, and then put the two pieces together.