- My Forums
- Tiger Rant
- LSU Recruiting
- SEC Rant
- Saints Talk
- Pelicans Talk
- More Sports Board
- Fantasy Sports
- Golf Board
- Soccer Board
- O-T Lounge
- Tech Board
- Home/Garden Board
- Outdoor Board
- Health/Fitness Board
- Movie/TV Board
- Book Board
- Music Board
- Political Talk
- Money Talk
- Fark Board
- Gaming Board
- Travel Board
- Food/Drink Board
- Ticket Exchange
- TD Help Board
Customize My Forums- View All Forums
- Show Left Links
- Topic Sort Options
- Trending Topics
- Recent Topics
- Active Topics
Started By
Message
re: Daily COVID Updated as of 11/2/20 8:00 PM
Posted on 4/14/20 at 4:07 pm to jeffsdad
Posted on 4/14/20 at 4:07 pm to jeffsdad
quote:
Today is going to be the worse day yet , crap.
It’s been exactly what the models have predicted so far. We are trending downwards in new cases, and are about to hit the peak in deaths soon. After this week it will only get better.
0
Posted on 4/14/20 at 4:10 pm to HailHailtoMichigan!
It's interesting that while the death numbers are having their normal Tuesday spike, the new case numbers are not and may show another decline.
Is this a result of the trailing deaths? Probably.
Is this a result of the trailing deaths? Probably.
Posted on 4/14/20 at 4:21 pm to LosLobos111
quote:
Pretty much only have reliable numbers wed-fri
Would you settle for Wednesday and Thursday?
Posted on 4/14/20 at 4:34 pm to Powerman
quote:Or a moving average
I mentioned a few days ago it might be better to start looking at weekly numbers to get an idea of the true trend if the daily numbers aren't reliable
Posted on 4/14/20 at 4:35 pm to Korkstand
is there an explanation as to why it is tuesdays, and not mondays, when the delayed weekend deaths are recorded?
Posted on 4/14/20 at 4:37 pm to Korkstand
What timeframe do you think would be appropriate for the moving average, 3 days is too short, I'm thinking 7 days would be ideal?
Posted on 4/14/20 at 4:38 pm to HailHailtoMichigan!
Here's my guess. A patient dies over the weekend. Hospitals fill out the paperwork, paperwork goes to the medical examiner on Monday to do whatever they do. After approval, the paperwork goes back to the hospital at the end of the day. The hospital includes the death in their next reporting day...Tuesday.
My wife's mother died earlier this year on a Sunday. It took several days to get an the medical examiner to sign off due to systematic issues.
My wife's mother died earlier this year on a Sunday. It took several days to get an the medical examiner to sign off due to systematic issues.
This post was edited on 4/14/20 at 4:41 pm
Posted on 4/14/20 at 4:43 pm to Chromdome35
quote:I don't know, I would probably plot a few and see what looks best. I would guess 7 days too, but you might still get spikes and drops when last Tuesday rolls off and this Tuesday pops on. You could mostly eliminate the steps when a day rolls off with an exponential moving average, but that would just accentuate the most recent spike.
What timeframe do you think would be appropriate for the moving average, 3 days is too short, I'm thinking 7 days would be ideal?
I'm probably overthinking it and 7 days will work just fine.
Posted on 4/14/20 at 4:43 pm to Chromdome35
4 days
Posted on 4/14/20 at 4:45 pm to LSURussian
It may just be me, but the images on the first page aren't showing the data as of 3:30.
Posted on 4/14/20 at 4:47 pm to Korkstand
When you say exponential moving average do you mean a weighted moving average? By weighted average I mean you take the most recent result and multiply it by 7 (for a 7-day moving average) then take the day previous to that and multiple by 6, repeating until you get to a weight of 1. Total it all up then divide by 28 (7+6+5+4+3+2+1)?
Posted on 4/14/20 at 4:48 pm to imjustafatkid
Thanks Kid, fixed it!
Posted on 4/14/20 at 4:52 pm to HailHailtoMichigan!
Wheels of healthcare beauracacy turn very slow; particularly with deaths
Posted on 4/14/20 at 5:01 pm to Chromdome35
quote:FYI. For a moving average of times series data, there should be an equal amount of days on either side of the day calculated by the average, instead of the 7 days leading up to the final day.
What timeframe do you think would be appropriate for the moving average, 3 days is too short, I'm thinking 7 days would be ideal?
For example, if you do a 7 day moving average, the 4th day would be the midpoint, so then it should start on Day 4 and end on the 4th to last day (e.g., Saturday would be the midpoint through today).
In addition, if you’re concerned that 3 days is too few, then you might want to consider 5 days, since it would minimize some of their Saturday-Monday 3-day delay, while also being short enough to capture changes in the data. A 7 day moving average may be too long given the sample size and the rapidness of the changes.
This post was edited on 4/14/20 at 5:05 pm
Posted on 4/14/20 at 5:22 pm to Korkstand
This is a graph of the 7 day rolling average of New Cases and Deaths
I added a new column for 7 day rolling average deaths. It is on the right side of the deaths columns.
I added a new column for 7 day rolling average deaths. It is on the right side of the deaths columns.
This post was edited on 4/14/20 at 5:24 pm
Posted on 4/14/20 at 5:24 pm to Chromdome35
56 downvotes on the first page for saying deaths are a lagging indicator
Posted on 4/14/20 at 5:24 pm to Chromdome35
quote:Yeah that's what I meant.
When you say exponential moving average do you mean a weighted moving average?
Not sure how useful it would be though.
Posted on 4/14/20 at 5:25 pm to Chromdome35
I'm still impressed with your work and dedication to this project Chrome! 
Posted on 4/14/20 at 5:27 pm to LSURussian
quote:You don't think that's too short given there seems to be a 2-3 day lull every weekend?
4 days
Page 174 of 331
Popular
Back to top
Follow TigerDroppings for LSU Football News