Specify a Query in Search Cursor

[ZacharyHart]

Basically I'm trying to create a selection that satisfies a random criteria AND a query based on an attribute field. Here is some sample code I found on StackExchange:

def SelectRandomByCount (layer, count):
    import random
    layerCount = int (arcpy.GetCount_management (layer).getOutput (0))
    if layerCount < count:
        print "input count is greater than layer count"
        return
    oids = [oid for oid, in arcpy.da.SearchCursor (layer, "OID@")]
    oidFldName = arcpy.Describe (layer).OIDFieldName
    delimOidFld = arcpy.AddFieldDelimiters (layer, oidFldName)
    randOids = random.sample (oids, count)
    oidsStr = ", ".join (map (str, randOids))
    sql = "{0} IN ({1})".format (delimOidFld, oidsStr)
    arcpy.SelectLayerByAttribute_management (layer, "", sql)

What I need is this in conjunction with a query (from the same input later) where FIELD = Value (for a string field).

I have reviewed the ESRI help document related to Specifying a query in Python, but am having difficulty understanding how to implement this.

[JakeSkinner · Accepted solution]

Hi Zachary,

I would recommend creating a Feature Layer.  You can specify a query when doing so.  Ex:

arcpy.MakeFeatureLayer_management("C:/data/mexico.gdb/cities","cities_lyr", "POPULATION > 50000")‍

You can then pass the feature layer to the function to retrieved from StackExchange.

[ZacharyHart]

Jake, this is great and will totally get me there. To be clear, is there a way to build the sql query using a logical operator 'AND'? Here's an attempt but the query fails:

def SelectRandomByCount (layer, count):
    import random
    layerCount = int (arcpy.GetCount_management (layer).getOutput (0))
    if layerCount < count:
        print "input count is greater than layer count"
        return
    oids = [oid for oid, in arcpy.da.SearchCursor (layer, "OID@")]
    oidFldName = arcpy.Describe (layer).OIDFieldName
    delimOidFld = arcpy.AddFieldDelimiters (layer, oidFldName)
    TypeField = arcpy.AddFieldDelimiters (layer, "OS_TYPE")
    #TypeField = "OS_TYPE"
    Type = 'S3A'
    randOids = random.sample (oids, count)
    oidsStr = ", ".join (map (str, randOids))
    sql = """ {0} IN ({1}) AND {2} = '{3}' """.format (delimOidFld, oidsStr, TypeField, Type)
    print sql
    arcpy.SelectLayerByAttribute_management (layer, "", sql)

[EDIT: fixed...i just had to for

I ask this because what I'm actually looking to do this for the same field to satisfy 5 different queries...say for "FIELD" = 'Value1', then "FIELD" = 'Value2'...etc. I guess that would be a loop within a loop? If this qualifies as another question altogether, I'll mark yours as correct.

I'll hammer away at this, but do appreciate the insight!

[JakeSkinner]

Yes, you would want to do a loop.  Ex:

import arcpy
from arcpy import env
env.overwriteOutput = 1

values = ["Value1", "Value2", "Value3", "Value4", "Value5"]

for value in values:
    arcpy.MakeFeatureLayer_management(r"D:\Temp\Python\Test.gdb\Airports", "airports_layer", "FCC = '{0}'".format(value))
    SelectRandomByCount("airports_layer", 5)‍‍‍‍‍‍‍‍‍

[ZacharyHart]

Thanks Jake!