CSV Data Plugin and Import

Hello, everyone,

I have installed the CSV plugin and am trying to load data from a local file. The data source is recognized, including the first column as a time specification. The second column is always defined as a string, although the default is set to Number.
Because of this reason, the visualization as a bar chart does not work.

What could be wrong?

Thanks for your support

Did you consider using a transformation, such as this?

No, not via the subsequent transformation.

I found another solution last night and I think there is a small bug in the current plugin.

First I defined new channel names in the header. Then I set the correct values for field and type in the query and only then did I start the read import. The file was read correctly. Unfortunately, changing the type later will not be updated.

so what is the issue you are facing? I Am a bit confused, could be lack of caffeine

Does the csv file change the data in it often or is it static?

Well, I found a workaround for the problem. That’s good and maybe it’s because I’m a Grafana beginner. I had expected that a subsequent correction of the field type would update the import.

Unfortunately, the data is not entirely static. There is an update every month and I would like to show a comparison between the monthly current actual values and fixed target values. The monthly actual values can also come directly from InfluxDB
I had considered this solution for the static setpoints. Do you have any other and better solution ideas? In principle, it is a table month and default value that should fill a bar chart.

1 Like

Lots of unknowns here. What is the datasource of the csv? Why not get that data from the source?

the datasource is a simulation and a paperreport.

Alternatively, I can also use the test data db. However, the year of the static setpoint specification would have to be adjusted for the comparison display with the actual values. That would be easily possible in a CSV file on the drive via script.

In this isue I have one more question. Is it possible to filter out the end-of-month value for time and max. function value from a time series using, for example, transformation? (only one value pair per month)

→ In the screenshot, the yellow bar is to small and I believe because of the number of timesteps

Thanks for your support.

1 Like