From: Golden, Joshua J - (jgolden) Sent: Thursday, December 17, 2015 17:17 To: Marshall X Ma Cc: 'Fox, Peter'; 'Robert Hazen'; 'Daniel Hummer'; Downs, Robert T - (rdowns); 'Dimitri Sverjensky'; 'Chao Liu'; Anderson, Tanya Marie - (andersontmarie); 'Sophie Kolankowski'; 'Hao Zhong'; Morrison, Shaunna - (shaunnamm); 'Mike Meyer' Subject: Re: 3D Klee diagram Attachments: 3D-KLEE-30-elements-minerals-20151217.zip Dear all, I have attached an archive (.zip file) containing 30 .csv files. Each file name contains the element, the number of minerals with that element as of today (needed if we wish to convert to percent as the numbers can change as new minerals are added, e.g. 2 new minerals since yesterday), and today's date. The "element" "REE" includes lanthanides+Y as well as minerals in which the formula includes the generic "REE". Please discard the previous sample files for H as the was an issue with the REE numbers in those files. -Josh From: Marshall X Ma Sent: Thursday, December 17, 2015 2:28 PM To: 'Mike Meyer' Cc: 'Fox, Peter'; Golden, Joshua J - (jgolden); 'Robert Hazen'; 'Daniel Hummer'; Downs, Robert T - (rdowns); 'Dimitri Sverjensky'; 'Chao Liu'; Anderson, Tanya Marie - (andersontmarie); 'Sophie Kolankowski'; 'Hao Zhong'; Morrison, Shaunna - (shaunnamm) Subject: RE: 3D Klee diagram Thanks Mike for the comment. Yes, we can seek ways to realized that and provide a control panel on the user interface for the interactions. Thanks, Marshall Replied/forwarded by Xiaogang (Marshall) Ma TWC, RPI From: Mike Meyer [mailto:mike.meyer.geo@gmail.com] Sent: Thursday, December 17, 2015 12:45 To: Marshall X Ma Cc: Fox, Peter; Golden, Joshua J - (jgolden); Robert Hazen; Daniel Hummer; Downs, Robert T - (rdowns); Dimitri Sverjensky; Chao Liu; Anderson, Tanya Marie - (andersontmarie); Sophie Kolankowski; Hao Zhong; Morrison, Shaunna - (shaunnamm) Subject: Re: 3D Klee diagram Hi Marshall, I don;t know if it would be more/less work, or if its been discussed already, but it might work out that each element and/or interaction could have a transparency tab (or the option of controlling numerous ones via a 'master' tab) so that you could manually adjust transparency. Cheers, Mike On Thu, Dec 17, 2015 at 3:34 PM, Marshall X Ma wrote: Thanks Josh for the quick response, and Peter for the comments. Separate files in csv format will be ok. In my understanding each file will be the input of a matrix plane in the Z axis? You may just name the file with the Element name, like H.csv, Na.csv, etc. I can already forsee the steps to layer those planes up for a 3D cube matrix! Some further works are needed for adjusting the transparency of each cube, adding attributes, and more. Will have a short update with Bob and Dan on this at 1:45PM, central entrance to the Moscone South poster hall. Thanks, Marshall Replied/forwarded by Xiaogang (Marshall) Ma TWC, RPI -----Original Message----- From: Fox, Peter [mailto:pfox@cs.rpi.edu] Sent: Thursday, December 17, 2015 09:33 To: Golden, Joshua J - (jgolden) Cc: Robert Hazen; Ma, Xiaogang; Daniel Hummer; Downs, Robert T - (rdowns); Dimitri Sverjensky; Chao Liu; Mike Meyer; Anderson, Tanya Marie - (andersontmarie); Sophie Kolankowski; Hao Zhong; Morrison, Shaunna - (shaunnamm) Subject: Re: 3D Klee diagram Hi Josh, See below. > On 17 Dec 2015, at 9:14 , Golden, Joshua J - (jgolden) wrote: > > Dear All, > > Quick question for all. For the 3d analysis, should we use the counts or percents, so the number of minerals that contain all three elements or the percent of minerals that have element x, that also have y and z? Counts are better == raw data. > > Quick question for RPI team. Is it ok to provide the data for each > plane as a separate file? This is easier on my end. Also what file Yes, quite ok. We have to do some conversion for the graphing anyway, so the simpler the input data format, the better. A filename convention (ask Marshall) will help. > format do you prefer, .xlsx or .txt (tab delimited)? Also Re: the > first question, would the software be able to switch between comma separated values (csv). > displaying the counts and percents, or even calculate the percents from the counts given the number of minerals with element x? We will be able to add that in. Many thanks ---Peter. > > Once I receive the answers, it should not take me too long to put the files together, possibly less than an hour. > > Best regards, > > Josh > From: Shaunna Morrison > Sent: Wednesday, December 16, 2015 3:04 PM > To: Robert Hazen > Cc: Daniel Hummer; Marshall X Ma; Peter Fox; Downs, Robert T - > (rdowns); Golden, Joshua J - (jgolden); Dimitri Sverjensky; Chao Liu; > Mike Meyer; Anderson, Tanya Marie - (andersontmarie) > Subject: Re: 3D Klee diagram > > Bob et al, > > Yes, it is valid to treat Y + lanthanides as a single element, REE. > > In minerals, where you find one REE, you find others. They often occur together in groups of light rare earths (LREE) or heavy rare earths (HREE). > > LREE = La, Ce, Pr, Nd, Sm, Eu, Gd > HREE = Y, Tb, Dy, Ho, Er, Tm, Yb, Lu > > However, some minerals contain both LREE and HREE. > > With that said, it's probably best to start with treating all REE as a single element. > > > In writing this, it's made me think I should investigate REE relative abundances in greater detail, beyond light vs heavy, and see if there are other correlations. > > > Best regards, > > Shaunna > > > On Wed, Dec 16, 2015 at 12:15 PM, Robert Hazen wrote: > Hi Dan et al., > > I recommend that we generate a 30 x 30 x 30 matrix in which we look at coexisting elements X-Y-Z from the following list. This list includes all cations with at least 200 mineral species, and all anions with at least 100 species. If Dan can figure out the algorithm, each matrix element will represent the extent to which the coexisting 3 elements occur with greater or lesser frequency than what would be predicted by considering their crustal abundances alone. > > In this case the cube diagonal elements (X,X,X) are all 1. Any "special" plane of the cube (i.e., X,Y,Y), (X,X,Z), etc. will be the 2-D matrix calculated previously. I recommend that we start by arranging elements according to their period, not atomic number, as follows: > > Cations (grouped by period) > H > Na > K > Mg > Ca > Ba > Ti > V > Mn > Fe > Ni > Cu > Ag > Zn > B > Al > C > Si > Pb > REE [??] > U > > Anions (electronegativity; grouped by period) F (4.0) Cl (3.0) O (3.5) > S (2.5) Se (2.5) P (2.1) As (2.1) Sb (1.9) Bi (1.9) > > Ultimately, the 3D Klee diagram should have most matrix elements transparent, so we can highlight matrix elements that display either much greater or much lower occurrences than predicted by element abundances. Color coding, as Dan did on his poster (attached), would be a nice feature. I would expect that roughly cube-shaped subregions emerge from this exercise. > > One question is how we might treat REE. Shaunna, any thoughts? Is it valid to treat all lanthanides (+ Y??) as an element" in this process? > > Let me know what you think? This could be a great opportunity to develop and test this visualization tool. > > Best, > > Bob > > PS: Marshall, I think Dan and I can be outside 3000 Moscone West at 3 pm. I'll be there, in any case, if you want to discuss this. > > > > -- > Robert M. Hazen > Senior Staff Scientist, Geophysical Laboratory > > Executive Director, Deep Carbon Observatory > 5251 Broad Branch Road NW > Washington, DC 20015 > phone: > 202-478-8962 > > e-mail: > rhazen@ciw.edu > > > Personal web site: > http://hazen.gl.ciw.edu > DCO website: deepcarbon.net > Keck Deep-Time Project website: http://dtdi.carnegiescience.edu > > > > > -- > Shaunna Morrison > > Dept of Geosciences > 524 Gould-Simpson Building > University of Arizona > 1040 E 4th St. > Tucson, Arizona 85721-0077 > > Phone: (478) 737-5786 > > http://www.geo.arizona.edu/~shaunnamm/ -- ---- Mike Meyer, PhD Geophysical Laboratory Carnegie Institution for Science 5251 Broad Branch Rd. NW Washington, DC 20015-1305, USA https://dtdi.carnegiescience.edu/