Define a continuous time distribution:

Compute the mean of the distribution:

Simulate a time:

The distribution is left truncated at 0, which corresponds to midnight:

Define a discrete time distribution:

Compute the median of the distribution:

Simulate a time:

Simulate a list of times and visualize their distribution:

Define a continuous time distribution with mean at noon:

Compute the mean of the distribution:

Simulate a time:

Simulate a list of times and visualize their distribution:

Define a nonsymmetric distribution of times starting at 8am:

Simulate the instances:

Estimate TimeDistribution from the collection of times:

Since the data histogram is bell shaped, fit a normal distribution to the data:

Compute distribution functions:

Compute distribution moments:

The number of patients arriving at an emergency room has been recorded for each hour from 1am to midnight:

Create a nonparametric model of the data:

Use the model to create a distribution of times:

Simulate and visualize the sample:

Extract properties of a TimeDistribution:

List of properties:

Granularity of a discrete time distribution is given by the date scale unit:

Granularity of a continuous time distribution is always "Instant":

The input distribution must be a valid distribution:

Check if a distribution is valid:

Create a valid distribution:

The unit must be a valid time unit:

Check if a unit is a time unit:

For a time unit:

Or use KnownUnitQ with the second argument set to "Time":