How should type annotations look like for CSV to upload in the GUI?

InfluxDB 2
1

I want to import a CSV file which I created from a JSON file to my InfluxDB instance but apparently I am missing some type annotations…

This is the original JSON data:

{"time" : "2023-08-11 07:40:28", "model" : "Bresser-7in1", "id" : 34640, "temperature_C" : 20.700, "humidity" : 54, "wind_max_m_s" : 0.000, "wind_avg_m_s" : 0.000, "wind_dir_deg" : 138, "rain_mm" : 5.600, "light_klx" : 54.800, "mic" : "CRC"}
{"time" : "2023-08-11 07:40:53", "model" : "Bresser-7in1", "id" : 34640, "temperature_C" : 20.700, "humidity" : 52, "wind_max_m_s" : 0.000, "wind_avg_m_s" : 0.000, "wind_dir_deg" : 138, "rain_mm" : 5.600, "light_klx" : 55.200, "mic" : "CRC"}
{"time" : "2023-08-11 07:41:05", "model" : "Bresser-7in1", "id" : 34640, "temperature_C" : 20.800, "humidity" : 53, "wind_max_m_s" : 0.600, "wind_avg_m_s" : 0.600, "wind_dir_deg" : 138, "rain_mm" : 5.600, "light_klx" : 55.200, "mic" : "CRC"}
{"time" : "2023-08-11 07:41:30", "model" : "Bresser-7in1", "id" : 34640, "temperature_C" : 20.800, "humidity" : 54, "wind_max_m_s" : 0.700, "wind_avg_m_s" : 0.700, "wind_dir_deg" : 139, "rain_mm" :

which I convert with Python to

#datatype measurement,tag,field,field,field,field,field,field,field,field,field,field,field,time
m,id,model,id,temperature_C,humidity,wind_max_m_s,wind_avg_m_s,wind_dir_deg,rain_mm,light_klx,mic,time,time
Bresser-7in1,34640,Bresser-7in1,34640,20.5,55,1.4,1.4,138,5.6,54.0,CRC,2023-08-11T07:38:10Z,2023-08-11T07:38:10Z
Bresser-7in1,34640,Bresser-7in1,34640,20.5,55,1.4,1.4,138,5.6,54.0,CRC,2023-08-11T07:38:23Z,2023-08-11T07:38:23Z
Bresser-7in1,34640,Bresser-7in1,34640,20.5,55,0.7,0.7,138,5.6,54.5,CRC,2023-08-11T07:38:48Z,2023-08-11T07:38:48Z
Bresser-7in1,34640,Bresser-7in1,34640,20.5,55,0.0,0.0,138,5.6,54.5,CRC,2023-08-11T07:39:00Z,2023-08-11T07:39:00Z

How should the type annotations look like?

I checked the docs but I could not come up with a working format. The dryrun from the command line worked fine: influx write dryrun -b weather -f ...

I tried things like this:

#datatype measurement,tag,field,long,double,long,double,double,long,double,double,string,dateTime:RFC3339,time
m,id,model,id,temperature_C,humidity,wind_max_m_s,wind_avg_m_s,wind_dir_deg,rain_mm,light_klx,mic,time,time
Bresser-7in1,34640,Bresser-7in1,34640,20.5,55,1.4,1.4,138,5.6,54.0,CRC,2023-08-11T07:38:10Z,2023-08-11T07:38:10Z
2

Hello @tamasgal,
Here are some resources:

The best way to import a CSV imo is to use pandas and write a dataframe with the python client because i agree getting the annotations right is such a freakin pain.