My parents live in TX. As I was exploring the data to create my "COVID Danger Warning" python script for them(or for me so I can warn them if necessary) I found that the Texas Department of State Health Services(TDSHS) is publishing the deaths each day even if they haven't updated them. I also noticed that they continue to update deaths as far back( and probably further) as 2 weeks ago. So I wanted to check NYT deaths data against TDSHS. I found that NYT is a little slower to update the deaths, but that their updates create curves that seem to more accurately describe the current rate of deaths. Note: this data is particularly for the deaths in the county that my parents live in.
Essentially the data from the TDSHS always has the last 3-5 days as having the same number of deaths. What this means is that if you are looking at TDSHS data to see if the rate of deaths is currently increasing or decreasing it always looks like its staying the same because the last 2-5 days are always the same. These days don't get fully updated until about two weeks later as deaths from each day trickle in overtime.
I tried to get the data to have months so you can see the timeline better, but alas, I'm still learning how to do this with Matplotlib. What you are looking at in the graph is all the dates from March 7th until August 17th. You can see that the TDSHS data shows more deaths than NYT but their curve is flat at the top.
Below are some examples of the differences in the Texas data over time. The last date in the picture shows which day the data you see was published.
