====Reading Historical Data====

ARDI uses its own query language named AQL to ask for information.

The **GETHISTORY** function in AQL gets the history for one or more points of data. But these individual data points will often have different time-stamps, update rates and be a bit scattered. Trying to figure out exactly what the state of the system was at any given time can be challenging.

Using Pandas and the [[AQLQuery_Execute_DF|Execute_DF]] function, Python will process your data to provide it as a single, [[complete dataframe]].

===Creating a Query===

Once you're connected to the ARDI server, you'll need to create an AQLQuery

<code>
query = ardi.AQLQuery(server)
</code>

===Executing the Query===

You execute the query using the [[AQLQuery_Execute_DF|Execute_DF]] function, passing the actual AQL query you'd like to run.

The response comes back as a Pandas dataframe, with each column being named after each asset/property combination found.

<code>
frame = query.Execute_DF("'Closest Support' PROPERTY ALLPOINTS {\"range\": \"6 hours\"} GETHISTORY")
</code>

===Process the Response===

The index of your dataframe is //time//. You'll have one column for each point in the query, each named for the asset and property found. For example an asset called //Main Pump// with a //Temperature// measurement on it would be called "Main Pump.Temperature" in the resulting dataframe. 

===Complete Code===
<code>
import ardi
import sys

#Setup ARDI Connection
svr = ardi.Server("demo.optrix.com.au","long")

#Connect to ARDI Server
connected = False
try:
    connected = svr.Connect()
except:
    pass

if connected == False:
    print("Unable to Connect to ARDI Server")
    sys.exit(-1)

query = ardi.AQLQuery(svr)
data = query.Execute_DF("'Closest Support' PROPERTY ALLPOINTS {\"range\": \"6 hours\"} GETHISTORY")

print(str(data))
</code>