A Letter of Recommendation for Nan Xiao | /en/2014/11/lor-nan-xiao/

yihui 2022-12-16 23:59:26

A Letter of Recommendation for Nan Xiao

https://yihui.org/en/2014/11/lor-nan-xiao/

5 Comments

giscus-bot 2022-12-16 23:59:27

Guest *Nan Xiao* @ 2014-11-19 12:25:28 originally posted:

Thanks Yihui. I am totally moved by this sincere LoR.

I couldn't agree more with you for the importance of hacking skills today, especially in disciplines like statistics. Statistics is actually the exploration process to understand the mechanism of something, and in such process, we either need to hack the sample space (frequentists), or hack the parameter space (Bayesians). I think that is part of the reason why there is a lot of fun in it. The hacking process to make theoretical methods feasible and even making data-driven products, is of much fun, too. To some extent, they are all connected.

If I may, I want to suggest an additional disclaimer for the gif: No animals or guys were harmed during "the boosting". :D

@nanxstats

giscus-bot 2022-12-16 23:59:32

Guest *doe doe* @ 2015-02-15 16:24:48 originally posted:

Nan Xiao, you rock. Your Rcpi package is the first Proteocheometrics R package and I really really like it. Its so cool. From retrieving protein sequences and compound smiles to calculating fingerprint descriptors to z-scale descriptors to cross-terms. You got everything on Rcpi package. You save my life. And your web server that you are involved is so cool mate. I can just calculate morgan descriptors which was used to be only available in rdkit. And also, I like the Protr websever. you rocks mate. Going to date with R now. Take care.

giscus-bot 2022-12-16 23:59:33

Guest *Nan Xiao* @ 2015-02-16 02:50:36 originally posted:

Thanks! I am very happy to see that you found the packages and web applications useful. Rcpi and protr/protrweb have been published on Bioinformatics, and you can read them here: http://bioinformatics.oxfordjournals.org/content/31/2/279 and http://bioinformatics.oxfordjournals.org/content/early/2015/01/23/bioinformatics.btv042 Please feel free to contact me or send pull requests on GitHub if you have any questions when using them.

@nanxstats

giscus-bot 2022-12-16 23:59:34

Guest *doe doe* @ 2015-02-16 09:44:06 originally posted:

Cool man,I will do pull request when I am ready. Please take it easy on me. I will do pull request on compound and compound cross terms. This is what I steal from your Rcpi tutorial. haha
'tensorprod' - column-by-column (pseudo)-tensor product type interactions, d has p1 × p2 columns.

the getCPI works perfectly on Compound and Protein cross-terms but not for Compound and Compound or Protein and Protein because p1 and p2 is the same for self cross-terms. Right now, I am manually doing it on excel, and i tell you, its a pain man. So, I will have to find refuge in R. Use your getCPI function from Rcpi as a steping stone to get self cross terms. Maris and my boss said self cross terms is N * N-1 * 0.5. I will try to do it first and let you know how it goes. Kk.

giscus-bot 2022-12-16 23:59:28

Guest *liyun* @ 2014-11-22 04:37:21 originally posted:

Jiangtang says Nan is on the market so I thought it was the "job" market. Oh well, so it is actually the pre-Ph.D market. Best of luck.

@cloudly

giscus-bot 2022-12-16 23:59:29

Guest *John Vida* @ 2014-11-22 21:19:16 originally posted:

I am sure this will bring him great opportunities :) Very humble work Yihui! Of course, best of luck to Nan Xiao..!

giscus-bot 2022-12-16 23:59:30

Guest *Sunnysean* @ 2015-01-05 12:26:26 originally posted:

I cannot find variables that I build in rmd file when it is knited into html or pdf. Why?

giscus-bot 2022-12-16 23:59:31

Guest *aa* @ 2016-05-23 07:29:03 originally posted: