dboyliao

import numpy as np


class StridedIterator:
    """
    Reference:
       - https://github.com/python/cpython/blob/b2bf2bc1ece673d387341e06c8d3c2bc6e259747/Modules/itertoolsmodule.c#L2342
    """
    def __init__(self, begin, end, strides):
        self._idx_cnt = list(begin)

dboyliao

"""
Just for fun: https://www.facebook.com/groups/706762492694458/permalink/4414710665232937/
"""
import numpy as np


class Word:

    __map = {
        ("金", "金"): "鍂",

dboyliao

#!/usr/bin/env python3
# https://gist.github.com/dboyliao/ccb5970ccafd6d5c0420d39514cb0464
from pprint import pprint

import numpy as np


class _Tuple:
    def __init__(self, *values):
        self.__values = values

dboyliao

import numpy as np

def distance_vec(X):
    # (x - y)^2 = x^2 + y^2 - 2xy
    prod = -2 * X.dot(X.T)
    sq = (X ** 2).sum(axis=-1)
    return sq + sq[:, None] + prod

dboyliao

import numpy as np
from itertools import combinations


def distance_loop(X):
    N, M = X.shape
    dist = np.zeros((N, N))
    for i, j in combinations(range(N), 2):
        xi = X[i]
        xj = X[j]

dboyliao

from utensor_cgen.backend.graph_lower import TensorAllocationPlanner
from utensor_cgen.frontend.tensorflow import GraphDefParser

parser = GraphDefParser(config={})
planner = TensorAllocationPlanner(config={})

ugraph = parser.parse('tests/deep_mlp/simple_mnist.pb', output_nodes=['y_pred'])
planner.apply(ugraph)
plan = ugraph.attributes[planner.KWARGS_NAMESCOPE].plan
total_size = ugraph.attributes[planner.KWARGS_NAMESCOPE].total_size

dboyliao

from ortools.sat.python import cp_model

model = cp_model.CpModel()
nonoverlap_map = {
    'w:0': [],
    'x:0': ['w:0'],
    'y:0': ['x:0', 'w:0'],
    'z:0': ['x:0', 'w:0', 'y:0'],
    'u:0': ['w:0', 'z:0'],
}

dboyliao

from ortools.sat.python import cp_model

model = cp_model.CpModel()
model.NewIntVar      # declare a variable of type int
model.NewBoolVar     # declare a variable of type bool
model.NewIntervalVar # declare an inclusive interval [a, b]
model.Add            # add constraint, such as x < 1 
model.AddNoOverlap   # add constraint for non overlapping intervals
# and more...

dboyliao

from ortools.sat.python import cp_model

# setup a model which represents the constraint optimization problem
model = cp_model.CpModel()
# declare variables
# x and y are two integer with range of 0 to 2
x = model.NewIntVar(0, 2, 'x')
y = model.NewIntVar(0, 2, 'y')
# add constraint: x+y <= 1
model.Add(x + y <= 1)

dboyliao

import torch

x = torch.tensor(0, requires_grad=True, dtype=torch.float64)
y = torch.tensor(0, requires_grad=True, dtype=torch.float64)
l = torch.tensor(0, requires_grad=True, dtype=torch.float64)
lr = 0.1

# min x^2+y^2 s.t x+y = 1
for i in range(100):
    L = x**2 + y**2 + l*(1-x-y)