# Selecting points within fishnet grid cell with criteria of at least 3 unique values using ArcGIS Desktop?

My goal is to create a raster from points with values based on at least 3 unique IDs within a cell. Cells that had fewer than 3 points with unique IDs will have value NA. See below for clarification. I looked at Selecting set number of points within polygon grid cells using QGIS? hoping to adapt it, but am still stuck.

I created a fishnet with certain cell size that I think may be reasonable. Now, I want to identify the number points within each cell with unique IDs (here, VesselID). Next, I want to assign NAs to my field of interest (here, AdjPounds) for points that do not fit the "minimum 3 unique IDs in the cell criteria". I thought about using the Intersect tool, but don't know how to add the criteria.

Can you help?

This is probably best done in Python, but I'm not familiar with that language yet.

More context: I will need to reiterate this process for several similar files and ultimately conduct raster math to get an average raster. I assume I will need to choose the same cell size when I create the rasters to avoid resampling to obtain the final "average" raster.

• What do you want in AdjPounds when not NA?
– Bera
Commented Aug 6, 2019 at 17:21
• I want AdjPounds to be sum of value in the cell. I assumed this would happen when I use the Points to Raster tool and choose Sum based on that field. Commented Aug 6, 2019 at 17:23
• So AdjPounds already has values and you want to Null out some of them based on your criteria? Which arcmap version do you have?
– Bera
Commented Aug 6, 2019 at 17:25
• 10.6. Also, trying intersect with fishnet and points and returning null output (because fishnet isn't a polygon?). Commented Aug 6, 2019 at 17:33

## 1 Answer

Intersect the points and fishnet polygons. Then one way is to use python and the pandas module to count unique point ids per fishnet id.

``````import arcpy
import pandas as pd

#Change these four lines to match your data
inter_output = r'C:\data.gdb\intersect123'
fishnet_id_field = 'FID_fishnet'
vessel_id_field = 'VesselID'
adjp_field = 'AdjPounds'

#Create a pandas dataframe by reading each record with da.SearchCursor
df = pd.DataFrame.from_records(data=arcpy.da.SearchCursor(inter_output, [fishnet_id_field, vessel_id_field]), columns=[fishnet_id_field, vessel_id_field])
#Count unique ids per fishnet
df = df.groupby(fishnet_id_field)[vessel_id_field].nunique().reset_index()
#Create a dictionary of the result with fishnet id as key and unique count as value
d = dict(zip(df[fishnet_id_field], df[vessel_id_field]))
#Update adjp_field using the dictionary
with arcpy.da.UpdateCursor(inter_output, [fishnet_id_field, adjp_field]) as cursor:
for row in cursor:
if d[row[0]]<3: #If less than 3 unique values
row[1] = None
cursor.updateRow(row)
``````

If you dont want to use python you should be able to:

1. Dissolve intersection by fishnet id and vessel id to remove duplicates
2. Count using summary statistics
3. Join results back to the points
4. Field calculate if count >2