On the list by probability sampling
- Input: a collection C of elements and a probability distribution p over C;
- Output: an element chosen at random from C according to p.
C has n elements, 1-n, the probability (p = (p [1], ..., p [n]). We only random.random () function, it will give us a uniform distribution [0,1 a float on]. the basic idea is to split [0,1] into n segments of length p [1] ... p [n] (Σ p [i] = 1). If uniformly in [0, 1] on the dot, the probability that it stopped in the i-th segment is p [i]. Thus can be implemented by random.random () function. Check stop place [0,1] which position, then return it that segment index where the python implemented as follows:
ref: https://scaron.info/blog/python-weighted-choice.html
On the list by probability sampling
import random
import collections
def weighted_choice(seq, weights):
assert len(weights) == len(seq)
assert abs(1. - sum(weights)) < 1e-6
x = random.random()
for i, elmt in enumerate(seq):
if x <= weights[i]:
return elmt
x -= weights[i]
def gen_weight_list(seq, gt_set, incline_ratio):
'''
:param seq:
:param gt_list:
:param incline_ratio:
:return:
seqe = [1,2,3,4,5]
gt_list = [3,5,7]
# incline_ratio = 0.9 # allocate this num of prob for random select gt's in sequence
'''
len_seq = len(seq)
# programmatic gen the prob list:
prob_list = []
gts_in_seq = [i for i in seq if i in gt_set]
len_gts_in_seq = len(gts_in_seq)
# item_ngt_in_seq = [i for i in seqe if i not in gt_list]
if len_gts_in_seq > 0:
prob_gt = incline_ratio/len_gts_in_seq
prob_ngt = (1-incline_ratio)/(len_seq - len_gts_in_seq)
else:
prob_gt = 0
prob_ngt = 1/len_seq
for idx in range(len_seq):
if seq[idx] in gts_in_seq:
# prob_list[idx] = prob_gt
prob_list.append(prob_gt)
else:
# prob_list[idx] = prob_ngt
prob_list.append(prob_ngt)
return prob_list
# add prob incline ratio for allocate heavier weight udr some conditions:
seqe = [1,2,3,4,5]
gt_set = set([3,5,7]) # conditions, if item in seq is also in this list, will be allocated higher weight.
inc_ratio = 0.8 # allocate this num of prob for random select gt's in sequence
prob = gen_weight_list(seqe, gt_set, inc_ratio)
select_seq = []
for i in range(10000):
select_seq.append(weighted_choice(seqe, prob))
# count the item in select_seq:
select_seq.sort(reverse=True) #optional?
item_Count = collections.Counter(select_seq)
print(item_Count)