A Case of Collaboration on the S22 Dataset

A Case of Collaboration On the S22 Dataset

This article is a case study for collaboration within the release version of ESTEST 1.0. Lets join Yuna & Gary as they try to collaborate using ESTEST to verify and study the S22 dataset with simulations run by Qbox and Quantum Espresso

Lets take a look at their email correspondence.

Yuna Gray 	 Thu, Sep 1, 2011 at 11:15 AM
Reply-To: Yuna Gray 
To: "gmyuan@ucdavis.edu" 
Hi Gary,

I've done the complete set of simulations of the S22 dataset using Quantum Espresso and I would like to verify these results against another code. Can you help me run the same dataset using Qbox so that we can compare?

Thanks.

Yuna
Gary Yuan 	 Thu, Sep 1, 2011 at 11:24 AM
To: Yuna Gray 
Hi Yuna,

Sure I can help you run these simulations using Qbox. I'm quite interested in how well we compare! Just one thing though, may I suggest that we collaborate using ESTEST for this project? I think the tools they provide will greatly help to expedite this work. You can sign up for an account on

http://estest.ucdavis.edu

Then just upload your documents!

Thanks.

Gary.

Yuna Gray then goes to upload her S22 Quantum Espresso documents to ESTEST.

Yuna Gray 	 Thu, Sep 1, 2011 at 12:01 PM
Reply-To: Yuna Gray 
To: Gary Yuan 
I've uploaded my 66 simulations to ESTEST. Here is a list of all the UUIDs that correspond to these documents

cd4e6f55-0126-4b37-a0bd-a9d045604943
59d6061f-ca39-46fa-857c-8debe8d2289c
90e23016-6a96-426f-9a3f-15e194da7f38
4e064837-1277-4c4c-a4ee-b47198c76de0
4fb7376b-7b52-4e80-b09c-a628eb1cbf6d
d8d206c1-8df7-40e4-a89e-d9d7f7cd352e
eff5056e-3dce-4bc1-a6f5-0590da7c6951
a8380b98-b0e3-4f07-a22b-cea4fa42d2ce
bf438f64-b19b-439b-bd63-0820b6156f62
6e078eaf-e25f-4bae-b94d-34eb70101a19
0727562b-a257-43bc-8f85-f37a1424ef6e
34ccd657-117c-4835-8137-ab6b80cf3c5c
5de92ebc-b00e-4002-a191-d320db9cd1a0
d20e95f1-4113-48f4-ab6a-495af58b4b5c
1a3064eb-b2d0-4f10-98ec-1bb53f4080d4
ad23b20f-b259-4c14-aab1-9abdebc41c8b
2245f95d-41aa-4102-a674-b6a879ffdc05
3aa3ffb0-0773-47e2-92e9-9a94ad58c4fa
197c13ef-b0ea-4ada-868a-d57a28a27b44
ce3b53f5-18d2-4fe3-9088-ebbfb08292cd
792ad4cf-8319-4c06-a78e-38d7afee46ed
1a377b46-a561-4450-be89-89ee127ee5e6
132c5486-c1ea-42de-af80-4778e76fafb4
5ec641e8-6a52-4346-b1b8-c97744c2e94d
18aff925-bb73-4627-94e3-8d325f919b2c
1bf88ad4-b34a-4ca4-977f-5ea54766d9e9
244634c0-34fe-4b01-bfde-2aca810a8dd0
e09ab6ba-45fa-43ec-8340-953b0d6d35bf
232c7b0d-4e3e-428a-b4c8-c8295117a9b9
01ad2236-74d5-409e-bd66-b200024fb5ef
8636b790-dc15-477a-9384-ceba894a4cdc
5f7e6e7e-897c-430c-9bc7-88055f4dc5e8
7f952f3b-d4d9-4bc3-b2a7-0d366da62f27
4db83529-a7a1-4477-bc1e-073af3279a12
b4fba004-b06e-4864-b513-5ffbfa50eadd
1e20e95d-937e-4b05-8c43-8a4a959afb2d
2930cf4a-3827-40a4-91cd-ed37ebf82a49
818fb502-f236-497c-8e66-e6bac00cd3d7
48427eb8-93a3-4bde-a366-9fce2f5e3717
fe23003f-3cde-443e-acf0-ee990fdfffbb
dabf8e3e-1e8d-43fd-a829- 7b0892c7fd42
c97837c9-b012-40c1-b5c1-dd9b4c6fbb2f
a3ae1f7c-3f76-4219-8398-5714c7bc0c54
6166aec5-98df-43b3-9e65-b554bad60b7e
6a3cc92a-e931-43ef-9e4a-6a9dac68fa72
7b18fba0-baef-405e-9450-48ed5c57cb17
29a5ca2f-8569-47d9-8f73-871483f52334
f8117b72-8d26-4188-87cc-d6525bc6a0dc
8f6ac689-b15c-433c-8d8e-c9fcf7ce20e9
338ae05f-9a21-4d9b-b082-0e1d72afdbdc
716028f4-6596-4b09-b5f0-23ad57bc1007
8b670adc-4361-4c5b-aaa8-eea7096ff0e0
c23f8b5f-f278-42f8-9e89-03160f2b4780
36da1c61-a962-434b-86c4-ae1e854a7650
cb94954e-2436-443e-92d6-4703a376c1f6
dfe7ff6a-b7d1-4f24-a5ef-3a479c3a2e77
3873a93d-c671-48ff-b932-ca1967721442
053ae08a-cd35-4801-937b-d54e84332185
59606b7e-2e66-4ce7-809c-9550f321e124
f733dec9-1702-48c2-867d-453327e7be3b
5235b312-4f7c-4747-a151-294b28ad749b
d06a e251-3c81-48e7-8143-7506a80e6feb
aff1b264-436b-4ff8-a63c-a65d868ea957
9d8a1f31-4594-40cf-b66b-2317f5147ea9
ded4643c-7d67-42ec-88b5-85a18400c78f
03ff43ae-a0d2-4d65-ab3e-303994e88c47

Is this okay?

Yuna.
Gary Yuan 	 Thu, Sep 1, 2011 at 12:06 PM
To: Yuna Gray 
Yuna, lets rename these documents with a common prefix so that we can easily search for them in ESTEST. May I suggest "S22/".

Gary.
[Quoted text hidden]

Yuna Gray then renames her documents as suggested.


Yuna Gray 	 Thu, Sep 1, 2011 at 12:12 PM
Reply-To: Yuna Gray 
To: Gary Yuan 
I've renamed the documents with the prefix you have suggested. What are the next steps?

Thanks.

Yuna
Gary Yuan 	 Thu, Sep 1, 2011 at 12:15 PM
To: Yuna Gray 
Next I'm going to take a look at your I/O data contained in the ESTEST UR XML documents and adapt them for running Qbox. Once I create the corresponding Qbox I/O and run them, I'll get back to you with my S22 data.

Thanks.

Gary.

Gary now has to get the S22 input data and convert the parameters from Quantum Espresso to Qbox format. This is quite the task when there are 66 simulations to consider! Fortunately ESTEST and some XSLT scripting can save him some time.

<?xml version="1.0"?>
<!--
Author: Gary Yuan
Date: 9/1/2011
File: ur2qbin_atomset.xslt
-->

<xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
<xsl:output method="text" />

<xsl:template match="/">
set cell <xsl:value-of select="/est_ur/est_data/atomset[@number='0']/unit_cell/@a" /> <xsl:value-of select="/est_ur/est_data/atomset[@number='0']/unit_cell/@b" /> <xsl:value-of select="/est_ur/est_data/atomset[@number='0']/unit_cell/@c" />
#
<xsl:for-each select="/est_ur/est_data/atomset[@number='0']/species">
species <xsl:value-of select="./@name" />
<xsl:choose>
<xsl:when test="./@symbol = 'H'"> http://fpmd.ucdavis.edu/potentials/H/H_HSCV_PBE-1.0.xml</xsl:when>
<xsl:when test="./@symbol = 'C'"> http://fpmd.ucdavis.edu/potentials/C/C_HSCV_PBE-1.0.xml</xsl:when>
<xsl:when test="./@symbol = 'N'"> http://fpmd.ucdavis.edu/potentials/N/N_HSCV_PBE-1.0.xml</xsl:when>
<xsl:when test="./@symbol = 'O'"> http://fpmd.ucdavis.edu/potentials/O/O_HSCV_PBE-1.0.xml</xsl:when>
</xsl:choose>
</xsl:for-each>
#

<xsl:for-each select="/est_ur/est_data/atomset[@number='0']/atom">
atom <xsl:value-of select="./@name" /> <xsl:value-of select="concat(' ',./@species)" /> <xsl:value-of select="concat(' ',normalize-space(./position))" />
</xsl:for-each>
</xsl:template>

</xsl:stylesheet>

Gary then fills in the input for Qbox and runs the simulations using Ranger @ UT. He then needs to upload the data to ESTEST; but he wants to do it efficiently and with the smallest upload archive possible.

Gary Yuan 	 Thu, Sep 1, 2011 at 3:00 PM
To: Yuna Gray 
Hi Yuna,

I've uploaded my S22 data run using Qbox. I've named these documents with the prefix string "yuangm/JSCH-S22-2005-IE-QBOX-PBE" that you can search for (and possibly filter by my user name 'yuangm'). You can then search on your own documents or (un)check them from your account manager and then use the "compare" transform xquery.

Thanks.

Gary.
[Quoted text hidden]

Yuna then goes to compare her QE/S22 documents to the ones Gary has uploaded on QB/S22.

Yuna Gray 	 Thu, Sep 1, 2011 at 3:09 PM
Reply-To: Yuna Gray 
To: Gary Yuan 
Hi Gary,

Thanks for doing this for me! I tried comparing as you suggested and I think we're getting good agreement but its kinda tedious to go through each of the S22 datasets and compare and compare... Is there a better way to look at an aggregate comparison?

Thanks.

Yuna.
Gary Yuan 	 Thu, Sep 1, 2011 at 3:29 PM
To: Yuna Gray 
Indeed there is! For the S22 dataset one of the derived values we can compute is the interaction energy, that is the difference in energy between the dimer and the two monomers. There is a plugin in ESTEST called interaction_energy that we can use to produce tables of the interaction energy that we can conveniently compare. I can show you how to use this plugin if you're having problems.

Thanks.

Gary.
[Quoted text hidden]

Gary shows how to run the interaction_energy plugin using 132 UUIDs in the input. This requires some special tactics because the cookie that normally conveys UUIDs saved to the plugins is limited to 99 values. Fortunately we can paste-in many more UUIDs directly into the plugin input form that transmits this data via POST.

Gary then sends the plain HTML output of the interaction_energy plugin to Yuna.

Gary Yuan 	 Thu, Sep 1, 2011 at 3:47 PM
To: Yuna Gray 
Hey Yuna,

I've attached a sample HTML output of the interaction_energy plugin that I've run with both our datasets! Hopefully this verifies your S22 simulation data using Qbox and Quantum Espresso.

Gary.
[Quoted text hidden]

Here is the attached HTML file.

Yuna Gray 	 Thu, Sep 1, 2011 at 3:51 PM
Reply-To: Yuna Gray 
To: Gary Yuan 
Thanks Gary!

I really like this output HTML document. Not only does it have the tables to compare the interaction energy calculated for the simulations using QE & Qbox, it has the links to the original data located on ESTEST.

Yuna.

Yuna seems happy! We hope this example has given you some ideas of how to collaborate within the ESTEST framework.