US numbers for 3/9/2020 and 3/10/2020 appear wrong

[matthewgapp]

The SUM for all US regions is 584 on 3/9/2020 and 1670 for 3/10/2020. These numbers should be 663 and 949 respectively according to other sources.

[tmeacham]

The naming convention for US locations changed on 3/10. Now it is reported at a state level only. Unfortunately, this messed up the time series data as for all days prior to 3/10 locations are not listed in this way, causing data duplication. if you filter out any location that contains a comma you you will get the correct number, but effectively have no way to do a time series for the united states as all cases suddenly appear on 3/10 in each state.

[matthewgapp]

Thanks. Is the 3/9 figure correct?

On Wed, Mar 11, 2020 at 9:59 AM Tom Meacham notifications@github.com wrote:

The naming convention for US locations changed on 3/10. Now it is reported at a state level only. Unfortunately, this messed up the time series data as for all days prior to 3/10 locations are not listed in this way, causing data duplication. if you filter out any location that contains a comma you you will get the correct number, but effectively have no way to do a time series for the united states as all cases suddenly appear on 3/10 in each state.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/CSSEGISandData/COVID-19/issues/495#issuecomment-597750051, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOYG3TQO4KFYMN6KAZKTDG3RG67PRANCNFSM4LF2F3NA .

[FrankSchiro]

So, should I be creating my own time series from raw data from this point on if tracking the US? Otherwise, will there be a fix to the time series file?

[williamlidata]

Older dates are wrong now. I see that the earlier dates have also aggregated to state level but the numbers don't match up, i.e. 1/30/2020, 1 case in Los Angeles, CA, but 0 in California.

[tmeacham]

Thanks. Is the 3/9 figure correct? … On Wed, Mar 11, 2020 at 9:59 AM Tom Meacham @.***> wrote: The naming convention for US locations changed on 3/10. Now it is reported at a state level only. Unfortunately, this messed up the time series data as for all days prior to 3/10 locations are not listed in this way, causing data duplication. if you filter out any location that contains a comma you you will get the correct number, but effectively have no way to do a time series for the united states as all cases suddenly appear on 3/10 in each state. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#495 (comment)>, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOYG3TQO4KFYMN6KAZKTDG3RG67PRANCNFSM4LF2F3NA .

Not if you filter out the old locations. Date before 3/10 will be messed up in the US unless they fix the data.

[FrankSchiro]

Obviously they have more granular data than offered on this git, do you know where we get the source data, so that we can compile with whatever naming conventions we want?

[matthewgapp]

Can you please fix the time series so that each date aggregates to the correct amount based on country? It's difficult to aggregate differently for different periods for different countries. Thank you!

On Wed, Mar 11, 2020 at 12:34 PM FrankSchiro notifications@github.com wrote:

Obviously they have more granular data than offered on this git, do you know where we get the source data, so that we can compile with whatever naming conventions we want?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/CSSEGISandData/COVID-19/issues/495#issuecomment-597827500, or unsubscribe https://github.com/notifications/unsubscribe-auth/AOYG3TVCX4XERM7PMMJY2VLRG7RURANCNFSM4LF2F3NA .

[Siphon880gh]

Each day, Los Angeles gives a number of cases and deaths at this website: http://publichealth.lacounty.gov/media/Coronavirus/

I was hoping it could be incorporated into the data here. Please make Los Angeles in California an exception to state-level only?

[Siphon880gh]

Each day, Los Angeles gives a number of cases and deaths at this website: http://publichealth.lacounty.gov/media/Coronavirus/

I was hoping it could be incorporated into the data here. Please make Los Angeles in California an exception to state-level only?

I am annoyed that I can't closely monitor growth in my county anymore so I created my own tracker app: https://wengindustry.com/tools/covid19/