Click the legend on the right to remove or display each curve; press play to see the progression of the curves over time. Hover over the graph to see options for zooming in, rescaling, and downloading. The legend displays the function parameters for functions defined below.

Click here to compare cases across countries.

Each data point corresponds to daily, public updates by Kuwait's Ministry of Health or Ministry of Information of the total number of confirmed COVID-19 cases in Kuwait. The colored curves represent the closest function to all of the cases up to that day.

Blue is exponential, \( \mbox{cases} = a_1 \cdot \exp(a_2 \cdot \mbox{ day})\)

Yellow is linear, \( \mbox{cases} = b_1 + b_2 \cdot \mbox{ day}\)

The linear and exponential models are fit with regression.

Observe the sometimes extreme, day-to-day changes in the best fit models, despite visually not changing much. The reason for this is twofold. The number of reported cases per day varies in an unknown fashion and cannot be captured by nice functions including linear, exponential, and even sigmoid curves. Each curve is the nearest function to the available data up to that day. This doesn't change the following day only if the number of confirmed cases the next day follows that exact same nice curve - even though it doesn't. In machine learning terminology, we overfit the data.

Data (used to be) updated in real-time with the most recent COVID-19 data.