Parametrize and loop KQL queries in JupyterLab - jupyter-lab

Question

My question is how to assign variables within a loop in KQL magic command in Jupyter lab. I refer to Microsoft's document on this subject and will base my question on the code given here:
https://learn.microsoft.com/en-us/azure/data-explorer/kqlmagic
1. First query below
%%kql
StormEvents
| summarize max(DamageProperty) by State
| order by max_DamageProperty desc
| limit 10
2. Second: Convert the resultant query to a dataframe and assign a variable to 'statefilter'
df = _kql_raw_result_.to_dataframe()
statefilter =df.loc[0].State
statefilter
3. This is where I would like to modify the above query and let statefilter have multiple variables (i.e. consist of different states):
df = _kql_raw_result_.to_dataframe()
statefilter =df.loc[0:3].State
statefilter
4. And finally I would like to run my kql query within a for loop for each of the variables within statefilter. This below syntax may not be correct but it can give an example for what I am looking for:
dfs = [] # an empty list to store dataframes
for state in statefilters:
%%kql
let _state = state;
StormEvents
| where State in (_state)
| do some operations here for that specific state
df = _kql_raw_result_.to_dataframe()
dfs.append(df) # store the df specific to state in the list
The reason why I am not querying all the desired states within the KQL query is to prevent resulting in really large query outcomes being assigned to dataframes. This is not for this sample StormEvents table which has a reasonable size but for my research data which consists of many sites and is really big. Therefore I would like to be able to run a KQL query/analysis for each site within a for loop and assign each site's query results to a dataframe. Please let me know if this is possible or perhaps there may other logical ways to do this within KQL...

Answer 1

There are few ways to do it.
The simplest is to refractor your %%kql cell magic to a %kql line magic.
Line magic can be embedded in python cell.
Other option is to: from Kqlmagic import kql
The Kqlmagic kql method, accept as a string a kql cell or line.
You can call kql from python.
Third way is to call the kql magic via the ipython method:
ip.run_cell_magic('kql', {your kql magic cell text})
You can call it from python.

Answer 2

Example of using the single line magic mentioned by Michael and a return statement that converted the result to JSON. Without the conversion to JSON I wasn't getting anything back.
def testKQL():
%kql DatabaseName | take 10000
return _kql_raw_result_.to_dataframe().to_json(orient='records')