influxdb3-explorer fails to write csv data with integer field values

[panda00hi]

I'm encountering an issue when using influxdb3-explorer to import CSV data into InfluxDB 3 core. Although the "Parse Data" and "Convert to Line Protocol" steps succeed, the "Write to Database" operation fails when a field column (e.g., WaterTemperature) contains integer values — even if they are represented as floats with zero decimals (like 17.00).

Steps to reproduce:

1. Upload a CSV file containing a numeric field (e.g., WaterTemperature) with values like 17, 18, or 17.00.
2. Parse and convert to Line Protocol successfully.
3. Attempt to write to the database.
4. The write fails.

chicago_water_sensors-format_test.csv

Expected behaviour:

1. Numeric values like 17 or 17.00 in CSV should be interpreted as floats when the target field is typed as float.
2. Alternatively, the system should auto-coerce integers to floats during write if schema allows it (to avoid user confusion).
3. At minimum, the frontend (explorer) should reflect potential type mismatches before attempting write.

Actual behaviour:
response error:
API response error: parsing failed for write_lp endpoint invalid column type for column 'WaterTemperature', expected iox::column_type::field::float, got iox::column_type::field::integer

Environment info:
influxdb3-explorer version: by docker from influxdata/influxdb3-ui:1.0.3

v1.0.3

influxdb 3 core version

Suggested Fixes

1. Update influxdb3-explorer CSV-to-LP logic to default numeric fields to float unless explicitly marked as integer.
2. Enhance InfluxDB 3’s write_lp endpoint to support implicit integer → float promotion for field values (if schema defines float).
3. Improve error feedback in UI to highlight type conflicts before write attempt.

This behavior creates a poor UX, as users are not aware of internal type handling differences between CSV parsing and database expectations.