akio tanakaのtumblr. @akio-tanaka - Tumblr Blog

eigenのcrossを使いたい場合、<Eigen/Geometry>をインクルードすること。コンパイルできてもリンク時にエラーが発生する。

you must include <Eigen/Geometry> to use cross. you will face link error without the including. (but no error while compiling)

#c++

3D PDFに関するご案内

フムフム

#3d

『あなたの人生の物語』(テッドチャン, 浅倉久志著)

Installing python-pptx

Easy to install lxml from package as one fo python-pptx dependeicies.

#python

#Python

subprocess.Popen

subprocess.Popen(cmd, shell=True)

How does shell in arguments work in Windows? When I used True, it can run but there was some errors during executing applications. And it's False, it can run normally...

#python

To give date and time as post-fix

'datetime' module is required to get and modify preciese time with os.path.splitext().

import os.path import datetime def give_date_and_time(file_path): root, ext = os.path.splitext(file_path) time_str = datetime.now().strftime('%Y%m%d_%H-%M-%S') return root + "_" + time_str + ext print(give_date_and_time('C:\test\input.txt'))

#python

Command line arguments in PyCharm

You need to give command line arguments into ‘Script parameters:’ on 'Run/Debug Configurations' from 'Run' in menu bar then 'Edit Configurations...'

#python #pycharm

Sample of argparse

If it's difficult to understand 'argparse' to parse command line arguments in the official documentation, below sample is useful to confirm it with with runing, modifying and confirming the documentation.

Parsing result, args in the sample, is dictionary with key name defined the 1st argument in parser.add_argument() and value input actually in command line arguments.

import argparse parser = argparse.ArgumentParser(description="program explanation when user input '-h', '--help' or invalid arguments") parser.add_argument('argument1', type=str, help='1st argument without any option. its position in arguments is arbitrary.') parser.add_argument('argumetn2', type=str, help='2nd argument without any option. its position in arguments in arbitrary.') parser.add_argument('-option1', metavar='option1(type is float)', type=float, help="to give option1 like '-option1 XXX'") parser.add_argument('-option2', metavar='option2(integer)', type=int, help="to give option2 like '-option2 YYY'") args = parser.parse_args() print(args)

#python

MLP 深層学習

* プーリング層 ** 最大プーリング *** ゼロパディング ** 平均プーリング ** LPプーリング ** ストライド設定可能 ** 学習パラメータなり * 正規化層 ** 白色化 *** 全訓練データが対象 ** 局所コントラスト正規化 *** 一枚画像の局所正方領域が対象 *** 減算正規化 **** 重み付き平均、正方領域中心が大きい *** 除算正規化 **** 重み付き分散 *** 単一キャネル *** 多チャネル ****全チャネルの平均・分散

We’ll show you how to train and optimize basic neural networks, convolutional neural networks, and long short term memory networks.

#deep learning #machine learning

京大発表。今後はSoC活用で処理を高速化

SoC活用で処理を高速化、の方向に進んでいるの

ありがたや

Assertions are very useful to catch programmer mistakes. Sadly, there is no mechanism to trigger an actual assert() in CUDA kernel code. Lung Sheng Chien showed me a simple way to create a assert f...

sm_20 or later required for printf in kernel function

#cuda

Judge NaN in C++

use the feature of NaN that returs true only when it is compared with itself.

generate

#include <iostream> #include <limits> int main() { std::cout ::quiet_NaN() ::signaling_NaN() #include <limits> template<typename t> bool IsNan(const T& v) { return v != v; } int main() { float nan = std::numeric_limits<float>::quiet_NaN(); float nan2 = std::numeric_limits<float>::signaling_NaN(); float nonNan = 0.1; std::cout </float></float></typename></limits></limits></iostream>

#c/c++

Trending Blogs

Recently Viewed Blogs

akio tanakaのtumblr.