test ylmcv

 

https://github.com/calico/ylmcv

HQ first cloned this on Apple Macbook, but failed to create the conda environment. 

HQ then tried on ECS323lambda, and was able to create the conda environment after correcting a typo. 

However, HQ ran into a pickle load error. After googling, HQ installed git-lfs 

  sudo apt-get install git-lfs

HQ then re-cloned the repo, and then the test script runs. 

(base) hqin@ECS323Lambda:~/ylmcv$ conda activate ylmcv

(ylmcv) hqin@ECS323Lambda:~/ylmcv$ ls

checkpoint  dataset.py environment.yml  examples  img_aug.py  layers.py  LICENSE  main_predict.py  model  README.md  transforms.py  util.py

(ylmcv) hqin@ECS323Lambda:~/ylmcv$ bash examples/count.sh 

using model: model.model_count

number of test samples 5 

Testing round 1

Testing round 2

Testing round 3

(ylmcv) hqin@ECS323Lambda:~/ylmcv$ cd examples/

(ylmcv) hqin@ECS323Lambda:~/ylmcv/examples$ ls output/

cnt_output.pkl

(ylmcv) hqin@ECS323Lambda:~/ylmcv/examples$ cd ..

(ylmcv) hqin@ECS323Lambda:~/ylmcv$ bash examples/

count.sh       data/          event.sh       output/        test_list.txt  

(ylmcv) hqin@ECS323Lambda:~/ylmcv$ bash examples/event.sh 

using model: model.model_event01

number of test samples 5 

Testing round 1

Testing round 2

Testing round 3

using model: model.model_event02

number of test samples 5 

Testing round 1

Testing round 2

Testing round 3

(ylmcv) hqin@ECS323Lambda:~/ylmcv$ 

Machine Learning with PyTorch and Scikit-Learn

 

https://github.com/rasbt/machine-learning-book

cpsc4900 Ch5 cloud sofware, part 1

 == pre-class to do: 

calendar email invitation: including guests;  NA 

socrative questions.  done

update Canvas course materials, update learning objectives. assignments as needed.   NA

* make sure grades are NOT shown on Canvas. 

 learning objectives:  ??

== In-class to do: 

For ECS Room 220, podium computer use channel 4 in Extron controller. 

restart computer to make sure ZOOM works. make sure Bluetooth mouse are turned off. 

clean up desktop space, calendars: 

ZOOM, live transcript (start video recording).  Turn the computer speaker on. 

* finished Ch4, socrative, Ch5. stopped at "15 list advantages and disadvantages of multi-instance databases"

*go over team docs on product roadmap and software architecture

let student work on Product backlog and Spring 1 plan. 

cpsc4900 2022 Feb 22

== pre-class to do: 

calendar email invitation: including guests;  NA 

socrative questions.  done

update Canvas course materials, update learning objectives. assignments as needed.   NA

* make sure grades are NOT shown on Canvas. 

 learning objectives:  ??

== In-class to do: 

For ECS Room 220, podium computer use channel 4 in Extron controller. 

restart computer to make sure ZOOM works. make sure Bluetooth mouse are turned off. 

clean up desktop space, calendars: 

ZOOM, live transcript (start video recording).  Turn the computer speaker on. 

* announcement, assignments and exams. 

* finished Ch4, socrative, started Ch5. 

*go over team doc

let student work on Project software architecture and road map. 

Database, platform, server, open source, development tools

Interpretable Deep Learning under Fire

Providing explanations for deep neural network (DNN) models is crucial for their use in security-sensitive domains. A plethora of interpretation models have been proposed to help users understand the inner workings of DNNs: how does a DNN arrive at a specific decision for a given input? The improved interpretability is believed to offer a sense of security by involving human in the decision-making process. Yet, due to its data-driven nature, the interpretability itself is potentially susceptible to malicious manipulations, about which little is known thus far

https://www.usenix.org/system/files/sec20spring_zhang_prepub.pdf

interpretable machine learning Christo[h Molnar

 

https://christophm.github.io/interpretable-ml-book/

beta values of DNA methylation

 

Beta values (β) are the estimate of methylation level using the ratio of intensities between methylated and unmethylated alleles. β are between 0 and 1 with 0 being unmethylated and 1 fully methylated.

Reference: 

https://www.biostars.org/p/182108/#:~:text=Beta%20values%20(%CE%B2)%20are%20the,unmethylated%20and%201%20fully%20methylated. 

Information Flow in Deep Neural Networks

 Information Flow in Deep Neural Networks

https://arxiv.org/abs/2202.06749

Estimating Information Flow in Deep Neural Networks

https://arxiv.org/abs/1810.05728

deep learning in bioinformations

 

https://github.com/liyu95/Deep_learning_examples

cpsc4900 Ch4, Feb 17, Thu

   == pre-class to do: 

calendar email invitation: including guests;  NA 

socrative questions.  done

update Canvas course materials, update learning objectives. assignments as needed.   NA

* make sure grades are NOT shown on Canvas. 

 learning objectives:  ??

== In-class to do: 

For ECS Room 220, podium computer use channel 4 in Extron controller. 

restart computer to make sure ZOOM works. make sure Bluetooth mouse are turned off. 

clean up desktop space, calendars: 

team formation, shared excel sheet, 

ZOOM, live transcript (start video recording).  Turn the computer speaker on. 

* announcement, assignments and exams. 

* start Ch4. socrative.   Ch4, start at "21"

*go over team doc on "user stories and feature identifications". ask students to make anonymous comments. 

Project software architeture: 

Database, platform, server, open source, development tools

Qin, short bio

 

Dr. Qin has 15 years of teaching and research experience in computational biology and data science. Dr. Qin currently works at the Department of Computer Science and Engineering and the Department of Biology, Geology and Environmental Science at the University of Tennessee at Chattanooga since 2016. Dr. Hong Qin worked at the Department of Biology at the Spelman College from 2009 to 2016 and the Tuskegee University from 2007 to 2009. Dr. Qin is a recipient of the William A Hinton Research training award from the American Society for Microbiology and a recipient of the NSF CAREER award

DBI Capacity

https://www.nsf.gov/pubs/2021/nsf21501/nsf21501.htm

Cyberinfrastructure Programmatic Area: https://www.nsf.gov/funding/pgm_summ.jsp?pims_id=505538&org=DBI&from=home

The Cyberinfrastructure Programmatic Area supports the implementation of, scaling of, or major improvement to cyberinfrastructure for biology that advances or transforms contemporary biology and that is broadly applicable to a wide range of researchers.

Biological Collections Programmatic Area: https://www.nsf.gov/funding/pgm_summ.jsp?pims_id=505541

The Biological Collections Programmatic Area supports major improvements to or digitization of biological collections and collection-based information, enabling the advancement of biological understanding in important research areas, and increasing the broader applicability of collections.

Biological Field Stations and Marine Laboratories Programmatic Area: https://www.nsf.gov/funding/pgm_summ.jsp?pims_id=5449&org=DBI

covid19 tweet sentiment on virus transmission

 average covid19 tweet sentiment are relatively unchanged from June to November 2020, but Rt peak during summer and winter. So, it would be hard to use a horizontal line to predict a changing curve with 2 peaks. 

Perhaps, additional features, such as tweet freqs, and retweets, can be used in for prediction. 

DNA methylation on aging and life span,

"DNA methylation regulates gene expression by recruiting proteins involved in gene repression or by inhibiting the binding of transcription factor(s) to DNA."

Typically, methylation turns genes “off” and demethylation turns genes “on.”

 DNA methylation on aging and life span, --> heretability, --> genotypes? 

missing heratability on aging with epistatsis, genetics, PPI, 

cpsc4900 ch4 software architecture finish, review stories and features

  == pre-class to do: 

calendar email invitation: including guests;  NA 

socrative questions.  done

update Canvas course materials, update learning objectives. assignments as needed.   NA

* make sure grades are NOT shown on Canvas. 

 learning objectives:  ??

== In-class to do: 

For ECS Room 220, podium computer use channel 4 in Extron controller. 

restart computer to make sure ZOOM works. make sure Bluetooth mouse are turned off. 

clean up desktop space, calendars: 

team formation, shared excel sheet, 

ZOOM, live transcript (start video recording).  Turn the computer speaker on. 

* announcement, assignments and exams. 

* start Ch4. socrative.   Ch4, start at "21"

*go over team doc on "user stories and feature identifications". ask students to make anonymous comments. 

screen cast and video editing tools

OBS

Davinci Resolve

1. Zoom offer a free account that allows for 45 minutes video and screen sharing. see

2. Microsoft Team 

3. Mac users can use QuickTime for screencast. 

4. ATOMI . https://atomisystems.com/download/ (Links to an external site.)

5.  OBS and other tools

NLP with transformer

natural language processing with transformers

https://transformersbook.com/  

https://github.com/nlp-with-transformers/notebooks

cpsc4900 Ch4, and user stories and feature identification

 == pre-class to do: 

calendar email invitation: including guests;  NA 

socrative questions.  done

update Canvas course materials, update learning objectives. assignments as needed.   NA

* make sure grades are NOT shown on Canvas. 

 learning objectives:  ??

== In-class to do: 

For ECS Room 220, podium computer use channel 4 in Extron controller. 

restart computer to make sure ZOOM works. make sure Bluetooth mouse are turned off. 

clean up desktop space, calendars: 

team formation, shared excel sheet, 

ZOOM, live transcript (start video recording).  Turn the computer speaker on. 

* announcement, assignments and exams. 

* start Ch4. socrative.   Ch4, stopped at "20 What are the advantages and disadvantages of using open-source software". Next lecture should start at Q21. 

*go over team doc on "user stories and feature identifications". ask students to make anonymous comments. 

Tyler in the back

Tayler at front

block chain education

 

block chain lab: 

https://github.com/BlockchainLabSU/SUBlockchainLabs

1D, 2D, 3D CNN

 

• In 1D CNN, kernel moves in 1 direction. Input and output data of 1D CNN is 2 dimensional. Mostly used on Time-Series data.
• In 2D CNN, kernel moves in 2 directions. Input and output data of 2D CNN is 3 dimensional. Mostly used on Image data.
• In 3D CNN, kernel moves in 3 directions. Input and output data of 3D CNN is 4 dimensional. Mostly used on 3D Image data (MRI, CT Scans, Video).

https://towardsdatascience.com/understanding-1d-and-3d-convolution-neural-network-keras-9d8f76e29610 

cpsc4900 Ch4

== pre-class to do: 

calendar email invitation: including guests;  NA 

socrative questions.  done

update Canvas course materials, update learning objectives. assignments as needed.   NA

* make sure grades are NOT shown on Canvas. 

 learning objectives:  ??

== In-class to do: 

For ECS Room 220, podium computer use channel 4 in Extron controller. 

restart computer to make sure ZOOM works. make sure Bluetooth mouse are turned off. 

clean up desktop space, calendars: 

team formation, shared excel sheet, 

ZOOM, live transcript (start video recording).  Turn the computer speaker on. 

* announcement, go over assignment, exams. Discussed due dates of project assignment. 

* start Ch4. socrative.   Ch4, stopped at "9 Suggest three ways of controlling architectural complexity."

*go over team doc on "personas and scenarios". ask student to make anonymous comments. 

ffmpeg -i movie.mov -vcodec copy -acodec cop out.mp4

ffmpeg -i movie.mov -vcodec copy -acodec cop out.mp4

running error on snow leopard.

gganimate and ffmpeg

=> Try to install gganimate in CoLab, it is unavailable. 

=> try in RStudio, 

File access error: 

Warning: file_renderer failed to copy frames to the destination directory  

file_renderer(dir = ".", prefix = "gganim_plot", overwrite = FALSE)

The frame sources:

/var/folders/bw/k6_tkc2142v1wh5r_1yhkqhc0000gp/T//RtmpIopU0P/1eec3399c50d/gganim_plot0001.png

I tried to change file permission in Finder, but my username is not allowed. 

The following codes worked: 

file_renderer(dir = "/Users/hqin/github/R-animation-sandbox/tmp/", prefix = "gganim_plot", overwrite = TRUE)
anim <- p + transition_time(year) +
  labs(title = "Year: {frame_time}")
gganimate::animate( anim, renderer=ffmpeg_renderer())

I tested and found out that I have ffmpeg installed. 

Reference: 

https://community.rstudio.com/t/warning-message-file-renderer-failed-to-copy-frames-to-the-destination-directory/45261/13

https://ottverse.com/create-video-from-images-using-ffmpeg/

conferences for covid19 tweets and 2020 election

IEEE/ACM international conference on advances in social networks analysis and mining, 10-13 November, 2022, Europe or virtual 

https://asonam.cpsc.ucalgary.ca/2022/dates.php#workshops-dates

8th international conference on human and social analytics, HUSO 2022, 22-26, May, Italy

https://www.iaria.org/conferences2022/HUSO22.html

cointegration project, Rt ~ Google workplace

2022-02-08:  

on ridgeside: Try cajo on Rt ~ G only. 

conda activate /home/hqin/.conda/envs/r2021nov

14:50 $bash commands_cajo-RtG.txt  #this ca.jo on Rt ~ workplace mobility. 

I had to close many sh tabs, and restart the shells. 

15:16 $bash commands_cajo-RtG.txt   #this runs for a while. 

outputfile:  filename = paste( "_autocajo_RtG", start, '-', end, ".csv", sep='')

Somehow, the output file has a wrong start and end labels. 

15:40-7, batch run from 1 to 3151, done in 7 minutes for Rt ~ G. 

 change output file name to _autocajo_RtG1-3151_ridgeside.csv

this is a full-time series co-integration analysis. 

2022-02-09 plot: 

cpsc 2800 in-person virtual box and linux tutorial

ex

 

facebook research and students awards

facebook research awards

https://research.facebook.com/research-awards/

phd https://research.facebook.com/fellowship/

  https://www.metacareers.com/students-and-grads/students/

CPSC4900 Ch3 questions, work on personas and scenarios

   == pre-class to do: 

calendar email invitation: including guests;  NA 

socrative questions.  done

update Canvas course materials, update learning objectives. assignments as needed.   NA

* make sure grades are NOT shown on Canvas. 

 learning objectives:  ??

== In-class to do: 

For ECS Room 220, podium computer use channel 4 in Extron controller. 

restart computer to make sure ZOOM works. make sure Bluetooth mouse are turned off. 

clean up desktop space, calendars: 

team formation, shared excel sheet, 

ZOOM, live transcript (start video recording).  Turn the computer speaker on. 

* start Ch3. socrative.   Ch3 questions. Start from Q20

* Announce Feb 8, I will be in Room 301, but I will start ZOOM breakout room to work on personas, and scenarios. 

skipped:  breakout rooms: personas, scenarios, user stories. This took most of time. 

skipped: in the main room, each team report discussion process. 

UTC wireless loggin problem on mac laptop

 Tech asked whether VirualBox or VMWare on my Apple laptop caused the wireless connection issues. 

In log-in, I typed username but I cannot type in the password. 

As suggested by the tech, I typed in the password first, and then my username, and it worked. 

cpsc4900 2022 02 01 Product vision, Ch3 continued

  == pre-class to do: 

calendar email invitation: including guests;  NA 

socrative questions.  done

update Canvas course materials, update learning objectives. assignments as needed.   NA

* make sure grades are NOT shown on Canvas. 

 learning objectives:  ??

== In-class to do: 

For ECS Room 220, podium computer use channel 4 in Extron controller. 

restart computer to make sure ZOOM works. make sure Bluetooth mouse are turned off. 

clean up desktop space, calendars: 

team formation, shared excel sheet, 

ZOOM, live transcript (start video recording).  Turn the computer speaker on. 

* breakout rooms, product vision (peer review), personas, scenarios, user stories. This took most of time. 

* in the main room, each team report discussion process. 

* start Ch3. socrative.  I did not have time for Ch3 questions. Start from Q20