# PANDAS-format not working
**URL:** https://community.cumulocity.com/t/pandas-format-not-working/1403
**Category:** Forum
**Tags:** cumulocity, datahub
**Created:** [June 24, 2024, 1:38pm UTC](https://community.cumulocity.com/t/pandas-format-not-working/1403 "2024-06-24T13:38:24Z")
**Posts on this page:** 5
**Page:** 1
### Author:  [@David.Richter](https://community.cumulocity.com/u/David.Richter)
#### Post date: [June 24, 2024, 1:38pm UTC](https://community.cumulocity.com/t/pandas-format-not-working/1403/1 "2024-06-24T13:38:24Z")
After I failed with my attemt to established a JDBC [Connection to DateHub](https://community.cumulocity.com/t/jdbc-connection-to-datahub/7696/1) I try to use the datehub Microservice instead.
I try to use the datahub microservice to query the data.I use the “High-performance API” which have the format-option (as the doc say [Cumulocity IoT DataHub - OpenAPI Specification](https://cumulocity.com/api/datahub/#operation/postPerformanceSQLApiResource)). In reality the parameter do nothing. I send as `format: "PANDAS"` but the result I get back look like the default forma, what is awkward to parse.
My Query:
`{"sql":"SELECT * FROM txxxxxxxxDataLake.\"pw-hmi\".txxxxxxxx.\"measurements_State_Offline\" WHERE source=241481197242","limit":1,"format":"PANDAS"}`
The result have this format:
```auto
{
"columns":[
"dir0",
"dir1",
"dir2",
"dir3",
"id",
"creationTime",
"creationTimeOffset",
"creationTimeWithOffset",
"YEAR",
"MONTH",
"DAY",
"time",
"timeOffset",
"timeWithOffset",
"source",
"type",
"Activity_Idle.series_times.unit",
"Activity_Idle.series_times.value"
],
"data":[
[
"chunk_P1_598_618_1701302400000_1701561600000_46597706",
"0_2023",
"0_11",
"0_30",
"1343464808",
"2023-11-30T05:53:44.532",
0,
"2023-11-30T05:53:44.532",
"2023",
"11",
"30",
"2023-11-30T05:53:13.966",
0,
"2023-11-30T05:53:13.966",
"881342014107",
"Activity_Idle",
"sec",
147915.0
]
]
}
```
The PANDAS-Format from the documentation looks much cleaner and better to parse with normal object mapper.
---
### Author:  [@Vachaspati\_Diwevedi2](https://community.cumulocity.com/u/Vachaspati_Diwevedi2)
#### Post date: [June 25, 2024, 10:03am UTC](https://community.cumulocity.com/t/pandas-format-not-working/1403/2 "2024-06-25T10:03:51Z")
Hi @David.Richter As per the API documentation it providing the right result set.
If the format is “DREMIO” the result set must have “Schema”, “rows” & “rowcount” like below
{
- “schema”:[
- “rows”:[
- “rowCount”: 5
}
If format is “PANDAS” result set must have columns and data only
{
- “columns”:[
- “data”:[
}
In your API response you are getting the columns and list of data (records).
Is your requirement is to do some additional data formatting?
---
### Author:  [@David.Richter](https://community.cumulocity.com/u/David.Richter)
#### Post date: [June 26, 2024, 6:07am UTC](https://community.cumulocity.com/t/pandas-format-not-working/1403/3 "2024-06-26T06:07:07Z")
Hi @Vachaspati_Diwevedi2
as you can see in my example response I get the result in DREMIO-format even I send `"format":"PANDAS"` in my request.
I want use the PANDAS-format because its easier to use. When I see the example here [Cumulocity IoT DataHub - OpenAPI Specification](https://cumulocity.com/api/datahub/#operation/postPerformanceSQLApiResource) it look like I get a complete objects in data that I can parse with ObjectMapper in my code. And unter columns I get information about the data type. In The DREMIO-format I don’t have information about the data types.
I build now a complex code to working with the DREMIO-format but I don’t understand why I can’t use the better PANDAS-format.
---
### Author:  [@system](https://community.cumulocity.com/u/system)
#### Post date: [December 20, 2024, 11:12pm UTC](https://community.cumulocity.com/t/pandas-format-not-working/1403/4 "2024-12-20T23:12:12Z")
---
### Author:  [@system](https://community.cumulocity.com/u/system)
#### Post date: [June 18, 2025, 11:12pm UTC](https://community.cumulocity.com/t/pandas-format-not-working/1403/5 "2025-06-18T23:12:47Z")
This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.