ArthurDelannoyazerty

class TriangleIndexes(nn.Module):
    def __init__(self, adjacency_matrix):
        super().__init__()
        self.adjacency_matrix = adjacency_matrix

    def forward(self):
        # tensor of indexes of each neighbors of each nodes 
        nonzero = self.adjacency_matrix.nonzero()
        neighbors_one_indexes = nonzero.reshape(self.adjacency_matrix.shape[0],15,2)[:,:,1].clone()
        neighbors_two_indexes = neighbors_one_indexes[neighbors_one_indexes]        # Tensor for each 2 neighbors for each nodes (neighbors of neighbors)

ArthurDelannoyazerty

def sets_to_lists(nested_dict):
    for key, value in nested_dict.items():
        if isinstance(value, set):
            nested_dict[key] = list(value)
        elif isinstance(value, dict):
            sets_to_lists(value)
    return nested_dict

ArthurDelannoyazerty

def merge_dict(d1, d2):
    """Merge d2 into d1 with terminal values stored in sets."""
    for key, value in d2.items():
        if key in d1:
            if isinstance(value, dict) and isinstance(d1[key], dict):
                merge_dict(d1[key], value)
            else:
                if not isinstance(d1[key], set):
                    d1[key] = set([d1[key]])
                d1[key].add(value)

ArthurDelannoyazerty

import xmltodict
from bs4 import BeautifulSoup

with open(filepath, 'r') as f:
    file = f.read()
soup = BeautifulSoup(file, 'xml')
str_xml = str(soup)
dict_xml = xmltodict.parse(str_xml)

ArthurDelannoyazerty

def count_leaves_dict(element:dict) -> int:
    if isinstance(element, list): return len(element)
    if not isinstance(element, dict): raise Exception("Wrong data type")
    nb_leaves = 0
    for _, value in element.items():
        nb_leaves += count_leaves_dict(value)
    return nb_leaves

ArthurDelannoyazerty

d_sorted = dict(sorted(d.items(), key=lambda item: item[1], reverse=True))

ArthurDelannoyazerty

import matplotlib.pyplot as plt
import matplotlib.patches as patches
from random import random

def draw_path(options:list[int], choices:list[int], margin:int=1, lw:int=7.5, inch:int=5):
    fig, ax = plt.subplots()
    ax.set_aspect('equal')
    ax.axis('off')

    theta1_limit, theta2_limit = 0, 360

ArthurDelannoyazerty

ffmpeg -r 24 -i images/%01d.png -vcodec mpeg4 -y movie.mp4
          ‾‾    ‾‾‾‾‾‾‾‾‾‾‾‾‾‾‾         ‾‾‾‾‾    ‾‾‾‾‾‾‾‾‾
          ↑           ↑                   ↑          ↑
         FPS        images             format       name

ArthurDelannoyazerty

from torch.utils.tensorboard import SummaryWriter
writer = SummaryWriter()

def save_histograms(writer, step, model):
    def flatten_gru_weights(gru_layer):
        flattened_weights = []
        for param in gru_layer.parameters():
            flattened_weights.append(param.data.view(-1))
        return torch.cat(flattened_weights)
    

ArthurDelannoyazerty

import pathlib
import numpy as np
import cv2
from tensorboard.backend.event_processing import event_accumulator


event_filepath = 'runs/Apr19_11-02-02_RNSCWL0127/events.out.tfevents.1713517322.RNSCWL0127.15332.0'
output_dir = 'images/img_train'

event_acc = event_accumulator.EventAccumulator(event_filepath, size_guidance={'images': 0})