BUG:Option "ignore unknown" removes too many columns

[Kortenbach]

When using the CSV plugin in Grafana "explore" you can define a name and a datatype for each column.
Then, when you tick "ignore unknown" the columns that do not match the schema should disappear.
Instead every column disappears (sometimes except for the first).
When I use the query inspector I find that the datatype of each column is "string" instead of string, string, num, num.

This is my datasource:

tagName;comment;min;max
CHRONOS/PLC1/Reg400001;Modbus dataregister;15.0;30.1
CHRONOS/PLC1/Reg400002;Modbus dataregister;16.0;30.1
CHRONOS/PLC1/Reg400003;Modbus dataregister;17.0;30.2
CHRONOS/PLC1/Reg400004;Modbus dataregister;18.0;30.3
CHRONOS/PLC1/Reg400005;Modbus dataregister;19.0;30.4
CHRONOS/PLC1/Reg400006;Modbus dataregister;20.0;30.0 

(Edit)
I found that this has to do with the query editor in Grafana.
When I select the column type as "Number" then, sometimes, I see the 15.0, 16.0,... (strings) turn into 15, 16,... (Numbers). If that happens then everything works fine! (But I don't know how to trigger that yet)

(Edit 2)
I just succeeded to trigger the "number" field type again. This is the log from the Query history:
{"decimalSeparator":".","delimiter":";","header":true,"hide":false,"ignoreUnknown":false,"params":[["test","0"]],"path":"/AnaAlarms-csv","schema":[{"name":"tagName","type":"string"},{"name":"min","type":"string"}],"skipRows":0}
(not working)

{"decimalSeparator":".","delimiter":";","header":true,"hide":false,"ignoreUnknown":false,"params":[["test","0"]],"path":"/AnaAlarms-csv","schema":[{"name":"tagName","type":"string"},{"name":"min","type":"number"}],"skipRows":0}
(working)

(edit 3)
When building a dashboard with the CSV plugin I can look at and change the JSON that is sent. In this JSON I can look at the "min" field and change it's type from string to number. This is a work around.
However, this workaround (edit JSON) is not accessable from the alert-rule editor or the explorer.

(edit 4)
Final workaround: Started using the infinity plugin (supports CSV among others).

[Kortenbach]

Issue closed, but is it solved??

[zoltanbedi]

You solved it by using infinity datasource instead.

[Kortenbach]

OK
I would call that a work-around. :-)

[zoltanbedi]

I closed it because you found a workaround and because there wasn't much reaction from community in a year.