## reviewer #2
This tool starts a simple image server that lets you quickly flip through image files from a local directory using your web browser.
Also, it optionally shows a customizable form where you can take notes or answer questions about each image or set of images.
### Example uses:
- manual curation / review of sequencing data visualization images such as those generated by [REViewer](https://www.illumina.com/science/genomics-research/reviewer-visualizing-alignments-short-reads-long-repeat.html) for short tandem repeat loci
- machine learning training set creation
- reviewing a pile of photos
### Features:
- simple way to flip through many local image files using your web browser
- crawls a top-level directory to find .png, .jpeg, or .svg image files
- web interface: home page lists all images
- web interface: image pages show the image, an optional customizable form where you can take notes or answer questions about the image, next/previous page links, and optional other customizable info for context
- use subdirectories to group images. Any images found in the same subdirectory will be shown on the same page
-------
[![PyPI version](https://img.shields.io/pypi/v/reviewer2.svg?style=flat)](https://pypi.org/project/reviewer2/) [![Supported Python versions](https://img.shields.io/pypi/pyversions/reviewer2.svg)](https://shields.io/)
### Install:
```
python3 -m pip install reviewer2
```
### Run:
```
python3 -m reviewer2 # start server for all images in the current directory and subdirectories
```
Below are more examples. Run with `--help` to see all available options and their descriptions.
```
python3 -m reviewer2 -x temp -x keyword2 /path/dir-with-images # -x are keyword(s) of paths to skip and /path/dir-with-images is the top level dir to search instead of the current dir
python3 -m reviewer2 -t /path/user_responses.xls # change where user responses get saved (default: reviewer2_form_responses.tsv)
python3 -m reviewer2 -m /path/metadata.tsv # provide a metadata table
```
After the server is running, open your web browser to [http://localhost:8080](http://localhost:8080) to start reviewing images.
### Options:
- metadata table (`-m`)
It's often useful to add extra info to the image pages to help with review - such as image descriptions, quality scores, etc.
To enable this, there are several ways to specify arbitrary key-value pairs to add to specific image pages.
The 1st way is to put a file called `reviewer2_metadata.json` next to the image(s). All keys and values from this file
will appear on that image page. The 2nd way is to use `-m` to pass in a metadata table
(`.tsv` or `.xls`) with a `Path` column + arbitrary other columns. If the `Path` value matches the relative directory containing
the image(s), entries from that row will be added to this image page.
Since the keys and values are treated as html, they can be used to add more complex info - such as
colors, text formatting, <img ..> tags with images from other web pages, iframes containing entire sections of external pages, etc.
- responses table (`-t`)
As you fill in the forms at the top of the image pages, the responses are written to this table. If you later restart reviewer2 with the same `-t`, it will reload previous responses. You can also optionally use this table to provide additional columns to display - sometimes this can be more convenient than using `-m`.
*Default*: `reviewer2_form_responses.tsv`
- custom form schema (`--form-schema-json`)
If you'd like to use non-default questions in the image page forms, you can specify the path or url of a .json file containing a custom form schema. For examples of the expected format see [main/form_schema_examples](https://github.com/broadinstitute/reviewer2/tree/main/form_schema_examples)
- config file (`~/.reviewer2_config`)
Most settings that can be provided on the command line can also be set via this YAML config file instead. For example:
*~/.reviewer2_config*
```
form-schema-json: /path/to/my-schema.json
hide-metadata-on-home-page: true
host: 127.0.0.1
port: 8080
```
For more details, run:
```
python3 -m reviewer2 --help
```
### Comparing reviews:
If 2 or more people review the same images, the reviews can be compared using the `compare_form_response_tables` script.
For example:
```
python3 -m compare_form_response_tables egor_reviewer2_form_responses.tsv ben_reviewer2_form_responses.tsv -o combined_responses.tsv -s1 egor -s2 ben
```
This script will print concordance stats, and output a `combined_responses.tsv` table that contains one image per row as well as each person's review of the image.
To see the full list of args and descriptions, run `python3 -m compare_form_response_tables --help`
### Development:
To create a local dev instance, run
```
git clone git@github.com:broadinstitute/reviewer2.git
cd reviewer2
# start server in dev mode so it reloads code on change
python3 -m reviewer2 /path/dir-with-images --dev-mode
```
挣扎的蓝藻
- 粉丝: 14w+
- 资源: 15万+
最新资源
- FeiQ.rar 局域网内通信服务软件
- 172.16.100.195
- 光储并网simulink仿真模型,直流微电网 光伏系统采用扰动观察法是实现mppt控制,储能可由单独蓄电池构成,也可由蓄电池和超级电容构成的混合储能系统,并采用lpf进行功率分配 并网采用pq控制
- python编写微信读取smart200plc的数据发送给微信联系人
- 光储并网VSG系统Matlab simulink仿真模型,附参考文献 系统前级直流部分包括光伏阵列、变器、储能系统和双向dcdc变器,后级交流子系统包括逆变器LC滤波器,交流负载 光储并网VSG系
- file_241223_024438_84523.pdf
- 质子交膜燃料电池PEMFC Matlab simulink滑模控制模型,过氧比控制,温度控制,阴,阳极气压控制
- IMG20241223015444.jpg
- 模块化多电平变器(MMC),本模型为三相MMC整流器 控制策略:双闭环控制、桥臂电压均衡控制、模块电压均衡控制、环流抑制控制策略、载波移相调制,可供参考学习使用,默认发2020b版本及以上
- Delphi 12 控件之FlashAV FFMPEG VCL Player For Delphi v7.0 for D10-D11 Full Source.7z
资源上传下载、课程学习等过程中有任何疑问或建议,欢迎提出宝贵意见哦~我们会及时处理!
点击此处反馈