DB structure and Google Sheets disappeared

[alessandroNa]

I think that, for subsequent and in-depth analysis, it is more convenient and efficient to have a file structure based on this way:

| Data | CountryRegion | ProvinceState | WHORegion | Type | NrType |

in the field Type must be indicated in the typeof data, that is confirmed, or deaths or recovered.

Is it possible to create a file like the one suggested on google sheets and accessible by the main business intelligence software? I use qliksense cloud.

[bjarkimg]

But then we miss the fun and excitement of transpose and merge the data sets... 😄

[alessandroNa]

@bjarkimg it is true but this takes time away from the data analysis: P

[3verse]

hey @DataEnthusiast84 - I did something similar to expose as a tableau data connector:
CSV (merge of all daily reports + column for filename): http://35.211.243.122/pubdata/covid-19-all.csv
WDC for tableau (if you need): http://35.211.243.122/api/wdc/html/COVID-19_WDC.html

[alessandroNa]

@3verse thanks for sharing the data! I use Qlik Sense to analyse the set.
How did you get data merging?

[3verse]

Hi @DataEnthusiast84,
For this dataset I've set up a small cloud server instance to regularly pull updates from this repository. After pulling, a bash script triggers to merge all the CSVs from the daily report folder together into one file including an extra column for the file name (the cvs I linked earlier). I didn't bother removing the repeated header names from the CSV as this is done in the WDC stage, another script that parses the CSV and translates into a Tableau compatible format, also fixing the inconsistent double quotes with some province/region values.

I'm more than happy to share everything about this process if you need

[alessandroNa]

@3verse

I really thank you for the information.
I'm still at the beginning on github.
In another discussion I "pushed" to have a file organized according to DB (and not individual files) since Qlik allows me to connect to a file via a connector. In truth I still have problems on the connector too.

I'm working on the DB file ... maybe when I finish I share it here ... I also pointed out that there are discrepancies (at 19/2) on the number of deaths. The data shows 2007 while 2014 is reported on the dashboard.

Study in progress ..

[3verse]

Uhm I thought that was just due to the data on here not being promptly updated to what the dashboard is showing.

For reference, I'm looking at the following one for live updates:
https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.html#/bda7594740fd40299423467b48e9ecf6

This is the viz I've built on Tableau Public based on the data my cronjob is pulling and merging regularly from this repo:
https://public.tableau.com/profile/salvatore1114#!/vizhome/wdctest/COVID-19StatusUpdate