Iteration algorithm over geometric attributes of shapefile with arcpyscripting

Question

I am trying to create a separate skyline shapefile for every feature ID I have in my point file by splitting them into multiple files. The code works fine, except splitting them into multiple shapefiles takes extremely long time. I have 65000 points in the shapefile. I have attached a photo of how I split my files. Once I split them, I point to each shapefile using the below code.

Splitting so many points currently takes about 25-30 hours on my computer. I use QGIS for this because I do not have advanced license for arcpy.Split_analysis tool to work with ArcMap.

Is there a better way than splitting the point file every time?

import arcpy, glob
# reading all the shapefiles
arcpy.env.workspace = "C:\\centrelinepoints"
files = glob.glob("C:\\centrelinepoints\*.shp")      


#running the skyline function on each point file separately.

for file in files:

    ## creating a unique name extension for each skyline shape output file based on its feature ID.    
    name = file                
    parts = name.split('_')      
    newname = parts[5]
    final = newname.split('.')
    FID = final[0]


    ## definiing the output layer location and name based on their feature ID.
    output_layer = "H:\\skylineshapes\Skyline_" + FID + ".shp"
    output_angle = "H:\\skylineangles\Skyline_" + FID + ".dbf"

    arcpy.Skyline_3d(file, output_layer, inSurface)

    arcpy.SkylineGraph_3d(file, output_layer, "0", "ADDITIONAL_FIELDS", output_angle)

Answer 1

Not guaranteeing that this is the best, but it made 1000 single point shapefiles in 171 seconds, which would still be just over 3 hours for 65000, but better than 25 hours!

import arcpy
fc = r'C:\junk\split\thousand_points.shp'
count = 0
with arcpy.da.SearchCursor(fc,'*') as cursor:
    for row in cursor:
        arcpy.CreateFeatureclass_management(r'C:\junk\split','shp_' + str(count),'POINT',fc)
        iCur = arcpy.da.InsertCursor(r'C:\junk\split\shp_' + str(count) + '.shp',cursor.fields)
        iCur.insertRow(row)
        del iCur
        print (count)
        count += 1