Pickups 拾遗

发表于 更新于 阅读次数：

每天进步一点点.

2020

18 Dec

R 语言stepAIC函数报错error in eval (predvars data env)，原因是因变量未指定。使用.省略因变量，可采用如下写法：

1
2
lm1 = lm(mpg ~ ., mtcars)
step1 = stepAIC(lm1, direction = "both", trace = FALSE)

另见passing-lm-result-to-stepaic-works-in-script-fails-inside-function.

2021

23 Mar Ctrl + Alt + M 快捷键占用

原因是与GeForce Experience快捷键冲突, 见What is preventing ctrl-alt-m from reaching IntelliJ on Windows 10?.

06 June 修改.gitignore后刷新

1
2
3
git rm -r --cached .
git add .
git commit -m '.gitignore is now working'

另见

  • gist:4371622
  • Resync git repo with new .gitignore file

26 Aug 同步新电脑

Zotero

  1. Edit-Preferences-Sync-Settings 设置File Syncing到坚果云指定目录
  2. Edit-Preferences-Advanced-Files and Folders 设置Base directory到OneDrive指定目录
  3. 安装ZotFile, Tools-ZotFile Preferences-General Settings, 设置Custom Location到OneDrive指定目录, 设置subfolder参数为/%c

镜像加速

  1. conda换源

Typora

语言为英文时无法选择PicGo, 需切换为中文.

Git

  1. Quickly set up GitHub SSH example. 运行ssh-keygen -o -t rsa -C "email@example.com", 然后把id_rsa.pub内所有内容复制进SSH keys / Add new.

  2. Best way to use multiple SSH private keys on one client [closed]. 例子如下:

    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    # github
    Host CarlossShi github.com
        HostName github.com
        User git
        IdentityFile ~/.ssh/github_CarlossShi

    # gitee
    Host CarlossShi gitee.com
        HostName gitee.com
        User git
        IdentityFile ~/.ssh/gitee_CarlossShi

  3. Git: “please tell me who you are” error. 例子如下:

    1
    2
    git config --global user.name "Carlos"
    git config --global user.email "ycshi2018@gmail.com"

30 Aug 在Python环境下安装Gurobi

How do I install Gurobi for Python?

31 Aug DP切换HDMI

三块屏幕, 主机Project选择Duplicate. 主屏幕用DP连接了主机, 两个HDMI分别连接Switch和PS4, 想要在三者间互相切换. 实验后发现

  1. DP连接主机后, 想要切换HDMI都会黑一下跳回DP
  2. 拔了DP, HDMI相互切换正常

参考一个显示器同时接HDMI和DP输出，可以选择切换输出吗？ - 地狱男爵的回答, 正确方式是Win+P把Project选为Second screen only, 然后盲切主屏幕.

17 Sep 微软拼音中文模式下使用英文标点

  1. win+I → 设备 (Devices) → 输入(Typing) → 多语言文本建议 (Multilingual text suggestions) → 关闭 (Off), 见微软拼音设置中文使用英文标点, 为什么还是出现中文标点???
  2. 若上述方法无效, 进入微软拼音设置, General → Use previous version of Microsoft Pinyin → On

https://docs.microsoft.com/en-us/windows/wsl/tutorials/gui-apps)

10 Oct WMT14 English-German预处理数据集

由于阅读Ghazvininejad2019MaskPredictPD代码facebookresearch/Mask-Predict需要, 要下载该数据集, 步骤如下:

  1. 新建fairseq/examples/translation/prepare-wmt14en2de.sh~

  2. 由于网络原因, 手动新建/orig/文件夹, 在其中下载上述代码中URLS内的5个文件

  3. 运行, 按照报错手动调整orig内的文件目录, 发现一堆报错

  4. ``` bash bash get_data.sh 

    1
    2
    3
    4
    5
    6

    2. Windows环境下手动运行`get_data`中的`preprocess.py`部分, 并把`--workers`参数改为1, 这是因为Windows不支持多线程, 具体如下: 

       ```bash
       python preprocess.py --source-lang en --target-lang ro --trainpref data/wmt16.en-ro/train --validpref data/wmt16.en-ro/valid --testpref data/wmt16.en-ro/test --destdir data-bin/wmt16.en-ro --joined-dictionary --workers 1 --nwordssrc 40000 --nwordstgt 40000
       python preprocess.py --source-lang ro --target-lang en --trainpref data/wmt16.en-ro/train --validpref data/wmt16.en-ro/valid --testpref data/wmt16.en-ro/test --destdir data-bin/wmt16.ro-en --joined-dictionary --workers 1 --nwordssrc 40000 --nwordstgt 40000

26 Oct 删除多版本Office软件

Fix Office installation errors

30 Oct .md转化为知乎适应格式

使用drmingdrmer/md2zhihu, 在博客目录创建.github/workflows/md2zhihu.yml如下, 其中pattern格式见Patterns to match file paths (避免一些不支持的Markdown格式在转换时报错).

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
name: md2zhihu
on: [push]
jobs:
  md2zhihu:
    runs-on: ubuntu-latest
    steps:
    - uses: actions/checkout@v2
    - uses: drmingdrmer/md2zhihu@v0.8
      env:
        GITHUB_USERNAME: ${{ github.repository_owner }}
        GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
      with:
        pattern: >
            source/_posts/*.md
            source/_posts/*.markdown
            !source/_posts/2020-11-22-Note on Hexo.md

14 Nov 治疗食指关节炎

How to use your keyboard to control the mouse pointer in Windows 10

How to Use a Keyboard to Click Instead of a Mouse

28 Nov Word中给空格加下划线

  • 用Tab和Ctrl+U加;
  • Home → Paragraph → Paragraph Settings → Tabs → Default tab stops, 设置一个Tab的长度.

30 Nov Word中解决空格换行问题

不知道为什么pandoc到处的docx文件中部分空格会换行, 导致导出的文档非常难看, 这些空格在show paragraph marks模式下显示为一个小黑点, 类似\(\ \cdot\ \), 解决办法是把它们全部替换为Non-Breaking Space, 快捷键为Ctrl+Shift+Space, 在上述模式下显示为一个空心源泉, 类似\(\ \circ \ \), 见Inserting a Non-Breaking Space.

三种方法见Microsoft Word Tricks: 3 Simple Ways to Keep Text Together.

3 Dec Jupyter Notebook父目录import

Importing modules from parent folder

24 Dec microsoft/recommenders

sequential_recsys_amazondataset

Tensorflow版本问题

报错cannot import name 'dynamic_rnn' from 'tensorflow.nn', 是版本不兼容问题. 解决方法为将sli_rec.py中from tensorflow.nn import dynamic_rnn中的tensorflow.nn改为tensorflow.compat.v1.nn.

重新看一眼原Notebook中输出的版本信息

1
2
3
4
5
print("System version: {}".format(sys.version))
print("Tensorflow version: {}".format(tf.__version__))
# System version: 3.7.11 (default, Jul 27 2021, 14:32:16) 
# [GCC 7.5.0]
# Tensorflow version: 2.6.1

然而不用%tensorflow_version 1.x会疯狂报错.

sli_rec.yaml

prepare_hparams中要输入的超参文件, 在Colab中的目录为/usr/local/lib/python3.7/dist-packages/recommenders/models/deeprec/config/sli_rec.yaml, 当前版本(dce8f71)没有, 可以到这里抄一个reco_utils/recommender/deeprec/config/sli_rec.yaml, 内容如下

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
#data
#data format:sequential model
data: 
    user_vocab : ./tests/resources/deeprec/slirec/user_vocab.pkl # the map file of user to id
    item_vocab : ./tests/resources/deeprec/slirec/item_vocab.pkl # the map file of item to id
    cate_vocab : ./tests/resources/deeprec/slirec/category_vocab.pkl # the map file of category to id

#model
model:
    method : classification # classification or regression
    model_type : sli_rec
    layer_sizes : [100, 64]  # layers' size of DNN. In this example, DNN has two layers, and each layer has 100 hidden nodes.
    att_fcn_layer_sizes : [80, 40]
    activation : [relu, relu] # activation function for DNN
    user_dropout: True 
    dropout : [0.3, 0.3]  #drop out values for DNN layer
    item_embedding_dim : 32 # the embedding dimension of items
    cate_embedding_dim : 8  # the embedding dimension of categories
    user_embedding_dim : 16 # the embedding dimension of users

#train
#init_method: normal,tnormal,uniform,he_normal,he_uniform,xavier_normal,xavier_uniform
train:
    init_method: tnormal  # method for initializing model parameters
    init_value : 0.01 # stddev values for initializing model parameters
    embed_l2 : 0.0001 # l2 regularization for embedding parameters
    embed_l1 : 0.0000 # l1 regularization for embedding parameters
    layer_l2 : 0.0001 # l2 regularization for hidden layer parameters
    layer_l1 : 0.0000 # l1 regularization for hidden layer parameters
    cross_l2 : 0.0000  # l2 regularization for cross layer parameters
    cross_l1 : 0.000   # l1 regularization for cross layer parameters
    learning_rate : 0.001
    loss : softmax    # pointwise: log_loss, cross_entropy_loss, square_loss  pairwise: softmax
    optimizer : lazyadam  # adam, adadelta, sgd, ftrl, gd, padagrad, pgd, rmsprop, lazyadam
    epochs : 50  #   number of epoch for training
    batch_size : 400  # batch size, should be constrained as an integer multiple of the number of (1 + train_num_ngs) when need_sample is True
    enable_BN : True  # whether to use batch normalization in hidden layers
    EARLY_STOP : 10 # the number of epoch that controls EARLY STOPPING
    max_seq_length : 50 # the maximum number of records in the history sequence
    hidden_size : 40 # the shape of hidden size used in RNN 
    attention_size : 40 # the shape of attention size
    need_sample: True # whether to perform dynamic negative sampling in mini-batch
    train_num_ngs: 4 # indicates how many negative instances followed by one positive instances if need_sample is True

#show info
#metric :'auc', 'logloss', 'group_auc'
info:
    show_step : 100   # print training information after a certain number of mini-batch 
    save_model: True   # whether to save modl
    save_epoch : 1    # if save_model is set to True, save the model every save_epoch.
    metrics : ['auc','logloss']  # metrics for evaluation.
    pairwise_metrics : ['mean_mrr', 'ndcg@2;4;6', "group_auc"]  # pairwise metrics for evaluation, available when pairwise comparisons are needed
    MODEL_DIR : ./tests/resources/deeprec/slirec/model/slirec_model/ # directory of saved models.
    SUMMARIES_DIR : ./tests/resources/deeprec/slirec/summary/slirec_summary/  # directory of saved summaries.
    write_tfevents : True  # whether to save summaries.

29 Dec Script file pip-script.py is not present.

  • 见Why Anaconda-script.py is not present ？ #762, easy_install.py可以在anaconda3\pkgs\setuptools-$\Lib\site-packages\setuptools\command中找到.

    1
    easy_install pip

  • 见Why Anaconda-script.py is not present ？ #762, 慢且有冲突.

    1
    conda install --force-reinstall pip

  • 见[Solved] Pycharm Use pip Error: Script file ‘D:-script.py‘ is not present.

    1
    python -m ensurepip --default-pip

注: 解决好后可能会收到提示要求你更新pip, 一通操作pip install --upgrade pip后发现pip又坏了.

2022

06 Feb pip install -e

Install a project in editable mode (i.e. setuptools “develop mode”) from a local project path or a VCS url.

What is the use case for pip install -e?

pip documentation v22.0.3 - #install-editable

09 Feb Inspect Variable

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
# [Printing Lists as Tabular Data](https://stackoverflow.com/a/26937531/12224183)
from tabulate import tabulate
def inspect(variable):
    try:
        result = []
        for md in variable.__dir__():
            if md[:2] == '__' and md[-2:] == '__':
                continue
            temp = getattr(variable, md)
            if isinstance(temp, list) or isinstance(temp, tuple):
                result.append([md, type(temp), len(temp), temp[:min(5, len(temp))]])
            elif isinstance(temp, dict):
                result.append([md, type(temp), len(temp), list(temp.items())[:min(5, len(temp))]])
            elif isinstance(temp, torch.Tensor):
                result.append([md, type(temp), temp.shape, temp[0:min(5, len(temp))]])
            elif isinstance(temp, str):  # has attr __len__
                result.append([md, type(temp), len(temp), temp])
            elif not hasattr(temp, 'len'):
                result.append([md, type(temp), '', temp])
        print(tabulate(result, headers=['Name', 'Type', 'Shape', 'Value']))
    except:
        print(md)

17 Feb Visual Studio Code with LaTeX

04 Mar 组合恒等式

今天好朋友给了个组合恒等式要证明, 发现了一些有意思的式子.

Show ∑n=0∞(2nn)xn=(1−4x)−1/2

Identity for convolution of central binomial coefficients: ∑k=0n(2kk)(2(n−k)n−k)=22n∑k=0n(2kk)(2(n−k)n−k)=22n

08 Mar Jamesellis51015/Multi-Agent-Path-Finding-with-Reinforcement-Learning Quick Start

今天帮一位老哥debug, 问题是运行训练时lanpa/tensorboardX的tensorboardX.SummaryWriter一直报错, 阅读文档tensorboardX后也没发现有啥问题, 弄了一个小时发现原来是events.out明明过长, 把xx_experiments.py中的name改短点就好了.

14 Mar 邮件问论文问题

今天看论文发现作者文章没写数据生成方式, 想起上次看Wiki看不懂也给作者发了封邮件, 每次都写的狗屁不通, 干脆整理一下.

To: %Corresponding author% Subject: %Paper title%

Dear Prof. %NAME%,

With a great interest I read your paper on %TOPIC%. Since I work in a similar area, I would like to %short presentation of what you want% [e.g., to compare my approach [1, 2, 3] to yours].

Do you have %your actual inquiry, detailed%?

Thank you very much in advance.

[1] Paper [2] Paper [3] Paper

Best regards, %Your name% -- %Your signature, including your institution, email, phone, and further ways to contact you%

– How to appropriately email a researcher to ask for details about their paper?

Dear Dr. [/Mr./Ms./Prof., try to know the appropriate title from the author list] [surname], I am interested in your paper [title] but I don't have access to it [you may briefly provide a reason why you need it/don't have access, if it may be interesting to them, e.g. what you are working on and where, collaborations/connections have been made that way], so can you please send it to me? Many thanks, [signature]

– How can I send requesting email to researcher for their research articles?

17h Mar 风扇管理

连着三天夜里跑模型自动关机了, 我觉得是GPU过热触发保护机制了, 因此找了个控制风扇速度的软件Argus Monitor, 见SpeedFan not detecting fans.

19 Mar JupyterLab

大三那时候看万门的课知道了Jupyter Notebook, 那时候不会用快捷键感觉真的好难用, 从一月打MathorCup大数据开始逐渐习惯用Jupyter Notebook做一些交互式的调试 (主要也就用abdm这几个快捷键), 今天发现没目录啊喂, 就发现了JupyterLab, 真好用!

可以在Setting → Advanced Setting Editor → JSON Setting Editor → User Preferences里设置自己喜欢的快捷键.

移动代码块的快捷键.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
{
    "shortcuts": [
        {
            <<other items you may have>>
        },
        {
            "command": "notebook:move-cell-up",
            "keys": [
                "Ctrl Shift ArrowUp"
            ],
            "selector": ".jp-Notebook:focus"
        },
        {
            "command": "notebook:move-cell-down",
            "keys": [
                "Ctrl Shift ArrowDown"
            ],
            "selector": ".jp-Notebook:focus"
        },
    ]
}

– David, How to move Jupyter notebook cells up/down using keyboard shortcut?

22 Mar jainor/ocst-problem

就问你能不能认真写Readme啊!

  • 代码两个make都漏了s, 要不是不知道make按tan看一下都发现不了, 应该改为 (其中ocstpin2可以改为src/test/TestInstances里的任意实例), 第一个cmake运行第一遍会出现CMake Warning: Manually-specified variables were not used by the project:InstanceOCSTP, 运行第二遍没问题.

    1
    2
    3
    cmake . -D InstanceOCSTP=ocstpin2
    make generate-instances
    make run-instances

  • scripts/python/runOCST.py中print语法不一致, 都改成Python3的就好了.

  • scripts/python/runOCST.py中把Queue改为queue.

  • 键入export LD_LIBRARY_PATH=src/lib/:lib/gurobi902, 否则会报错./src/bin/main: error while loading shared libraries: libproject_LIB.so: cannot open shared object file: No such file or directory和./src/bin/main: error while loading shared libraries: libgurobi90.so: cannot open shared object file: No such file or directory.

  • 运行make run-instances报错, 怀疑是多线程有问题, 已提issue.

    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    Exception in thread Thread-5:
    Traceback (most recent call last):
      File "/home/carlos/anaconda3/lib/python3.8/threading.py", line 932, in _bootstrap_inner
        self.run()
      File "/home/carlos/anaconda3/lib/python3.8/threading.py", line 870, in run
        self._target(*self._args, **self._kwargs)
      File "/home/carlos/Github/ocst-problem/scripts/python/runOCST.py", line 19, in parallelRun
        f=open(inputPath, "r")
    FileNotFoundError: [Errno 2] No such file or directory: 'Instances/sparseRsmallW/n20p0.2/in1'
    Thread  5  new task: 	 Instances/sparseRsmallW/n20p0.2/in3 Instances/sparseRsmallW/n20p0.2/pathbased/out3.csv pathbased

02 Apr conda Time Out

开始用组里的服务器了, 怎么装Pytorch这么慢😠?

先看看源有没有问题, 见ubuntu更换anaconda的镜像_泡泡里的月亮的博客, Anaconda 镜像使用帮助 - 清华大学开源软件镜像站. find /home/ -n .condarc查找后vim, 发现都加了💢.

使用conda config --show, 查看remote_read_timeout_secs (默认为66), 使用conda config --set remote_read_timeout_secs 1000修改.

怎么还是没速度👿, 究极解决办法 pytorch / packages/pytorch, 直接下载了安装.tar.bz2使用conda install --use-local ***.tar.bz2安装, 见Conda install 本地压缩包文件tar.bz2.

03 Apr 与两定圆相切的圆心轨迹

看了半天发现建模错了用不到, 不过这仍是个有趣的问题.

  • Given two tangent circles, find the two centers of a third that is tangent to both with given radius: 给定两圆是相切的
  • Find the center of a Circle tangent to two circles with a known radius, 给定两圆有重合
  • Given two circles in a plane, is there a simple equation that gives the locus of the centers of all circles that are tangent to both?: 很好的例子展示了2/4条轨迹
  • Locus of the centers of all circles tangent to two circles: 直观的轨迹动图
  • Find the position of a circle tangent to two other circles: 题设和图完全和我需求一致, 且要求是数学计算

30 Apr 卸载Office

帮室友装了两次Office, 总是安装报错Error Code 30094-1015, 猜想要把Office卸载干净, 用官方卸载工具最为省心, 见Uninstall Office from a PC, 下载地址为SaRA-officeUninstallFromPC. 然而还是继续报错, 由于这台电脑之前按照此方法装过Office猜想是激活后删除后存留了一些自动更新的模块无法停止, 最后还是选择装了学校的Office.

16 Jun Steam操作愿望单

  • [忤旭]《Steam》利用App ID添加游戏到Steam愿望单攻略

    1
    javascript:AddToWishlist(<APP_ID>, 'add_to_wishlist_area', 'add_to_wishlist_area_success', 'add_to_wishlist_area_fail', '1_5_9__407' );

  • Steam 愿望单无法移除锁区游戏的解决方法

  • Clash换区后使用桌面应用购买

21 Jun Github表情的使用

A First Look at Emoji Usage on GitHub: An Empirical Study: 用户如何在与开发相关的通信中使用表情符号的大规模实证研究

Beyond Textual Issues: Understanding the Usage and Impact of GitHub Reactions: 分析了开发人员对超过 250 万个问题和 970 万个问题评论提供的反应，以回答关于反应使用和采用的九个研究问题的广泛列表

The State of the Octoverse: communicating with emoji on GitHub: 社区如何使用表情符号表达自己的统计

17 Jul 遣散Teamviewer设备

由于新电脑坏了, 用了一阵子新电脑, 又要用到学校的台式机, Teamview登录上限到了, 要想办法替换电脑登录, 见Install Limit Reached - TeamViewer Community and Support, 实际上步骤应是Edit profile → Free devices, 一切一回我的Remainning switches已经是0了, 不知道之后怎么搞.

17 Jul 安装Windows Terminal

这东西大学捣鼓电脑时候装过, 很麻烦, 现在好像比以前方便多了, 在Microsoft Store里安装好后能直接检测到所有的可用Terminal, 比如Git, Powershell, Ubuntu等, 不用像以前一样一个一个手动配置. 之后考虑美化一下, 安装Oh My Posh.

Windows中在Powershell使用官方的三个方法, 发现最后一个即运行Set-ExecutionPolicy Bypass -Scope Process -Force; Invoke-Expression ((New-Object System.Net.WebClient).DownloadString('https://ohmyposh.dev/install.ps1'))可以安装, 见Installation-Windows

按照Installation-Fonts的cli方法直接安装发现连接不上, 又看到配置的字体是MesloLGM NF, 谷歌搜一下, 结果第一个是nerd-fonts/Meslo LG M Regular Nerd Font Complete Mono.ttf, 安装一下然后在settings → Defauts → Appearance → Font face里设置为MesloLGM Nerd Font Mono就可以了

按照Change your prompt在Powershell设置启动脚本, 发现权限不够, 则以管理员身份打开后运行set-executionpolicy remotesigned, 参见How to enable execution of PowerShell scripts?, 至此, Powershell的Oh My Posh已经配置完成.

而在WSL上首先按Installation-Linux的步骤安装, 随后参考Change your prompt在bash的设置方法即可, 具体为vim ~/.bashrc的末尾添加一行eval "$(oh-my-posh init bash)", 随后exec bash重载.

When using oh-my-posh in Windows and the WSL, know that you can share your theme with the WSL by pointing to a theme in your Windows user's home folder.

Inside the WSL, you can find your Windows user's home folder here: /mnt/c/Users/<WINDOWSUSERNAME>.

– Installation-Customize

设置主题时, Anacodna的Python环境不显示了, 看到了许多讨论

  • Help: Conda Environment Doesn't display · Discussion #1103: 指出使用conda config --set changeps1 false, 仍未解决
  • How to display current virtual environtment in python in oh-my-posh?: 指出sudo vim~/.bashrc中加入export VIRTUAL_ENV_DISABLE_PROMPT=1, 仍未解决
  • python segment is not showing current enviroment with miniconda on WSL2 #2323: 对theme的json文件中的python段进行了修改, 成功解决

下面来详细地整理一下. 由于尝试了一下午仍未解决, 最后意识到可能是和theme的配置有关, 我把配置放在了/mnt/c/Users/Carlos/.wt_theme.json, 下面给出两个版本的theme, 来回切换发现Windows Terminal是即时更新theme的

  1. Configuration-Overview中给出的极简版, 就一个路径

    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    12
    13
    14
    15
    16
    17
    {
      "$schema": "https://raw.githubusercontent.com/JanDeDobbeleer/oh-my-posh/main/themes/schema.json",
      "final_space": true,
      "blocks": [
        {
          "type": "prompt",
          "alignment": "left",
          "segments": [
            {
              "type": "path",
              "style": "powerline",
              "powerline_symbol": "\uE0B0",
              "foreground": "#ffffff",
              "background": "#61AFEF",
              "properties": {
                "style": "folder"
    }}]}]}

  2. jandedobbeleer.omp.json: 这就是一个让人恼火的版本, 需要修改三处地方

下面详细说一下三处改动, 即display_mode, fetch_virtual_env和template. 首先在json文件中查找python, 原始版本是这样的

1
2
3
4
5
6
7
8
9
10
11
12
{
    "background": "#FFDE57",
    "foreground": "#111111",
    "powerline_symbol": "\ue0b0",
    "properties": {
        "display_mode": "files",
        "fetch_virtual_env": false
    },
    "style": "powerline",
    "template": " \ue235 {{ if .Error }}{{ .Error }}{{ else }}{{ .Full }}{{ end }} ",
    "type": "python"
}

解释一下这三个要改动地方的含义

  • display_mode: string determines when the segment is displayed
    • always: the segment is always displayed
    • files: the segment is only displayed when *.py, *.ipynb, pyproject.toml, venv.bak, venv, or .venv files are present
    • environment: the segment is only displayed when a virtual env is present (default)
    • context: the segment is only displayed when either environment or files is active
  • fetch_virtual_env1: boolean - fetch the name of the virtualenv or not - defaults to true
  • template: Every segment has a template property to tweak the text that is displayed. Under the hood, this uses go's text/template feature extended with sprig and offers a few standard properties to work with. (see Configuration-Templates)

前两个改成default就行, template参照python segment is not showing current enviroment with miniconda on WSL2 #2323改为"template": " \ue235 {{ .Full }}{{ if .Venv }} {{ .Venv }}{{ end }} ", 改动后效果如下

image-20220718161922221

Appendix

Enable transparent backgrounds in Windows Terminal: 详细的步骤, 及troubleshooting

22 Aug 运行UCAS学位论文

尝试在VS Code中运行论文模板, 遇到如下问题:

  • Package ctable Error: You must load ctable after tikz.: 在table前加入tikz即可, 即\usepackage[*<super>,list,tikz,table,math*]{*Style/artratex*}
  • Tex/Chap_Guide.tex:214: LaTeX Error: Command \bfseries invalid in math mode.: 诸如$^{\citep{lamport1986document}}$应修改为$^\text{\citep{lamport1986document}}$, 见[Weird “

23 Aug Output Speaker to Microphone

Microphone input from speaker output回答得特别详细, 列举了一系列方法, 最后使用的是Voicemeeter. 具体方法为, 把virtual在输入输出都设为默认, 然后打开要播放的音频即可, 但此时该virtual设备是不会输出任何声音的. 若在腾讯会议中, 实际获取声音输出的是另一个设备, 可以将virtual输入的耳返输出到该设备.

02 Sep 清理C盘内存

  • SpaceSniffer, find lost disk space the easy way.
  • 移除pagefile.sys和hiberfil.sys获取C盘空间: swapfile.sys, hiberfil.sys
  • What is AppData Folder in Windows and How to Clean It Up?

29 Oct 制作动态Shields

到目标网站上的Network搜索想要的数据, 中文query使用$['净打点效率']格式填写, 否则报错unparseable jsonpath query, 见Unparseable jsonpath query error on dynamic json badge #5580.

进一步考虑在Typora中设置变量进行运算, 目前还没有此功能, 见Feature Request: Make use of YAML front matter variables #4191.