File indexing completed on 2026-07-26 08:22:17
0001
0002
0003
0004 import argparse
0005 import numpy
0006 import math
0007 import scipy.stats
0008 import random
0009 import collections
0010 import csv
0011
0012
0013 def neighbourhood(p, m):
0014 return [
0015 (x, y)
0016 for (x, y) in [
0017 (p[0] + 1, p[1]),
0018 (p[0] - 1, p[1]),
0019 (p[0] + 1, p[1] - 1),
0020 (p[0] - 1, p[1] - 1),
0021 (p[0] + 1, p[1] + 1),
0022 (p[0] - 1, p[1] + 1),
0023 (p[0], p[1] - 1),
0024 (p[0], p[1] + 1),
0025 ]
0026 if x >= 0 and y >= 0 and x < m and y < m
0027 ]
0028
0029
0030 def rand(d):
0031 return int(round(d.rvs()))
0032
0033
0034 def generate_file(name, Md, Hd, size, N):
0035 print("Generating file %s..." % name)
0036
0037 with open(name, "w") as f:
0038 h = 0
0039 w = csv.DictWriter(
0040 f,
0041 fieldnames=[
0042 "geometry_id",
0043 "measurement_id",
0044 "channel0",
0045 "channel1",
0046 "timestamp",
0047 "value",
0048 ],
0049 lineterminator="\n",
0050 )
0051 w.writeheader()
0052
0053 for m in range(N):
0054 acts = 0
0055 hits = rand(Md)
0056
0057 points = set()
0058
0059 for q in range(hits):
0060 cells = rand(Hd)
0061 center = (random.randint(0, size - 1), random.randint(0, size - 1))
0062
0063 cands = [center]
0064 seen = set()
0065
0066 for c in range(cells):
0067 if not cands:
0068 break
0069
0070 p = random.choice(cands)
0071
0072 if p not in points:
0073 w.writerow(
0074 {
0075 "geometry_id": m,
0076 "measurement_id": h,
0077 "channel0": p[0],
0078 "channel1": p[1],
0079 "timestamp": 0,
0080 "value": random.uniform(0.1, 1.0),
0081 }
0082 )
0083
0084 acts += 1
0085
0086 seen.add(p)
0087 cands += neighbourhood(p, size)
0088 cands = [x for x in cands if x not in seen and x not in points]
0089
0090 h += 1
0091 points |= seen
0092
0093
0094 if __name__ == "__main__":
0095 parser = argparse.ArgumentParser(description="Generate some CCL examples")
0096
0097 parser.add_argument(
0098 "-C",
0099 type=int,
0100 help="number of files to generate",
0101 )
0102 parser.add_argument(
0103 "-N",
0104 type=int,
0105 default=2500,
0106 help="number of modules in the detector",
0107 )
0108 parser.add_argument(
0109 "-S",
0110 type=int,
0111 default=655,
0112 help="size of each module (SxS)",
0113 )
0114
0115 parser.add_argument(
0116 "--Mm",
0117 type=float,
0118 default=1.65,
0119 help="mean hits per module",
0120 )
0121 parser.add_argument(
0122 "--Ms",
0123 type=float,
0124 default=0.95,
0125 help="scale of hits per module",
0126 )
0127
0128 parser.add_argument(
0129 "--Hm",
0130 type=float,
0131 default=1.80,
0132 help="mean cells per hit",
0133 )
0134 parser.add_argument(
0135 "--Hs",
0136 type=float,
0137 default=0.89,
0138 help="scale of cells per hit",
0139 )
0140
0141 parser.add_argument(
0142 "-o",
0143 type=str,
0144 default="ccl",
0145 help="output name",
0146 )
0147
0148 args = parser.parse_args()
0149
0150 hits_dist = scipy.stats.lognorm(args.Ms, scale=math.exp(args.Mm))
0151 cell_dist = scipy.stats.lognorm(args.Hs, scale=math.exp(args.Hm))
0152
0153 print("Number of modules: %d" % args.N)
0154 print("Module size: %dx%d" % (args.S, args.S))
0155 print("Total number of pixels: %d" % (args.N * args.S * args.S))
0156 print()
0157 print(
0158 "Dist parameters hits per module: %s(μ=%.2f, σ=%.2f)"
0159 % (hits_dist.dist.name, args.Mm, args.Ms)
0160 )
0161 print(
0162 "Dist parameters cells per hit: %s(μ=%.2f, σ=%.2f)"
0163 % (cell_dist.dist.name, args.Hm, args.Hs)
0164 )
0165 print()
0166 print("Expected hits per module: %.2f" % hits_dist.mean())
0167 print("Expected total hits: %.0f" % (args.N * hits_dist.mean()))
0168 print("Expected cells per hit: %.2f" % cell_dist.mean())
0169 print(
0170 "Expected cells per module: %.0f" % (cell_dist.mean() * hits_dist.mean())
0171 )
0172 print(
0173 "Expected total cells: %.0f"
0174 % (args.N * cell_dist.mean() * hits_dist.mean())
0175 )
0176 print(
0177 "Expected density: %.5f%%"
0178 % ((100 * hits_dist.mean() * cell_dist.mean()) / (args.S * args.S))
0179 )
0180 print()
0181
0182 if args.C is None:
0183 generate_file(("%s.csv" % args.o), hits_dist, cell_dist, args.S, args.N)
0184 else:
0185 for i in range(args.C):
0186 generate_file(
0187 "%s_%010d.csv" % (args.o, i), hits_dist, cell_dist, args.S, args.N
0188 )
