Distorted Imagery

A very quick post here but I just noticed this problem of one of my graphs being unintentionally distorted.

Here are the bird’s eye view of the cases in the Midwest. These sets of graphs allows me to see somewhat at a glance how each state in the Midwest is doing and I have included some tables of numbers at the bottom.

Image 1: Midwest bird’s eye view of cases

I’m going to focus on Minnesota. Note how since January 2021, Minnesota has been showing a U curve where cases started to go down sometime in January and then started to rise again sometime in March/April. The actual month of the rise is hard to see here but such exactitude is not my intent. I’m just noting that Minnesota is facing a real rise in cases and that it is a distinct U curve as compared to other states (excluding Michigan) that are having a flat or very minimal increase – so minimal that I question if the increase is really an increase or my imagination.

Here is another series of charts that I created as a set of “statistics” for a chosen state, in this example Minnesota. The first chart in the image below is showing the daily cases in Minnesota.

Image 2: State’s “statistics”

The U is very flat. Ah, this is a problem. What chart is “correct”? I’m not 100% sure but I am thinking the one that is showing a distinct U because the other one is actually squished – accidentally squished.

I did a series of experiments to check that the squished chart is actually like the first chart.

Image 3: Compare charts and play around with chart from State’s “statistics”

I copied the original chart from the Midwest’s bird’s eye view and placed it in another sheet in Power BI. That chart can be found in the upper right hand corner of the above image. The other original, that from the state “statistics” series, is located on the upper left hand corner.

The experiments are thus: For chart A in Image 3, I copied the original from the State’s “statistics” (upper left hand corner) and squished the chart by moving the right edge of the chart to the left. Presto, that one looks like the original from Image 1 (or the upper right hand chart). For chart B in Image 3, I copied chart A and took the bottom edge of the chart and squished the edge toward the top, and the original from State’s “statistics” reappeared. Finally, for chart C in image 3, I again took the original from State’s “statistics” and pulled the bottom edge of the chart downward to elongate the chart vertically, bringing back the U again (or like the original chart from the bird’s eye view series).

The big thing I have to be careful of is not squishing the charts too much to make them fit the page.