03:48:06 Sahil Rawa: Just joined now so is there anything that i have missed 03:48:29 Anirban Mandal: Replying to "Just joined now so i..." the real part is just getting started 03:49:40 Márcio Mateus Jr: Replying to "Just joined now so i..." Just presentations and general discussions about the workshop 03:55:20 Márcio Mateus Jr: whats the difference between RunG and RunH? Can I use the both datasets? 03:56:10 Márcio Mateus Jr: thanks! 03:56:43 Nandita.B: Will we get participation certificate ? 03:57:33 Matthew Bellis (he/him): Yes, we can provide participation certificates! 03:57:56 Caue Evangelista De Sousa: Exercise 1 and 2 done. I found on muon dataset with only muons(no electrons) and one electron dataset with both 03:58:03 Matthew Bellis (he/him): They will be emailed out as an image file, but it might take a few weeks to get them out. 03:58:07 Nandita.B: Replying to "Yes, we can provide ..." Thanks! 03:58:37 Jill Ziegler: Done with both parts! 04:00:53 Pedro Barbejat Lobo: I believe I'm done! 04:01:00 Jill Ziegler: The dataset I'm looking at is a muon set and has both muons and electrons. 04:02:19 João de Felipe Andrade: Done. 04:02:29 Sahil Rawa: done 04:05:46 Márcio Mateus Jr: In terms of the physics content, the mini and nano AOD format are equivalent? 04:10:48 Márcio Mateus Jr: I see, thanks! 04:12:15 Caue Evangelista De Sousa: In exercise 1 it's not so clear to find the pt, eta and phi in the miniAOD datasets, and the nanoAOD they just pop out when you open the variables list! 04:13:21 Julie Hogan: Yep, the 4-vector info is buried in this "LeafCandidate" parent class 04:13:57 Caue Evangelista De Sousa: In exercise 2 we can see that theres is much more information about the tracks in miniAOD than in the nanoAOD 04:15:12 Caue Evangelista De Sousa: In my master's I had an issue because I spent much time planning what track info to use only to found out it were no available on nanoAOD, sad but true 04:16:12 Márcio Mateus Jr: Replying to "In my master's I had..." how did you proceed? changed back to miniAOD? 04:16:23 Caue Evangelista De Sousa: Replying to "In my master's I had..." Exactly! 04:17:30 Caue Evangelista De Sousa: Replying to "In my master's I had..." I needed info about the hit pattern, point of closest approach, impact parameter, etc, to use on my discriminator ML model, so I had no choice but to accept that I had to come back to miniAOD 04:17:42 Julie Hogan: Replying to "In my master's I had..." MiniAOD analysis is not that bad in the end, but for sure the first level of analysis has to be done in C++ (let's say, making a file that is more like NanoAOD but has all the tracks you want) 04:19:15 Julie Hogan: Replying to "In my master's I had..." On Thursday we'll link back to some of the earlier workshops where we shared frameworks to help people analyze AOD or MiniAOD 04:19:24 Márcio Mateus Jr: Replying to "In my master's I had..." Oh, I see... interesting 04:23:23 Caue Evangelista De Sousa: Replying to "In my master's I had..." Wow, there is some nice information in this enhanced version of nanoAOD, like impact parameters 04:23:52 Caue Evangelista De Sousa: Replying to "In my master's I had..." also what seems the track quality 04:25:58 Annamalai S: Yes 04:26:04 Márcio Mateus Jr: So, one can always "enhance" the nanoaod with miniAOD data, or it just can be done for Particle Flow information? 04:27:53 vinny: Where can i find the recording of this video later? :) Thank you! 04:28:22 Julie Hogan: We will post them on the Indico page 04:30:09 Andrzej Konrad Siodmok: will the particle flow nano avaiable also for older data? 04:32:55 Julie Hogan: @Andrzej Konrad Siodmok. No. We created particle-flow Nano by adding a small thing to CMS's official production code for NanoAOD. For the older data we have some MiniAOD analysis frameworks available that users can run to get files like NanoAOD, including whatever features you like. But we don't have the production power right now to make them ourselves 04:38:33 vinny: Silly question, which dataset do you recommand that has enough data to find cool goodies in it 04:38:44 Márcio Mateus Jr: how can I filter the mass of the Z' ? 04:38:58 vinny: which dataset is comprehensive enough to work on? 04:40:36 Márcio Mateus Jr: it worked, thanks 04:41:32 vinny: Sorry one more question, which is the rawest dataset? is there one that is raw sensor data? 04:42:02 vinny: rawest/usable data 04:42:04 Ramakrishnan GK: Which kind of Dataset is mainly focused for finding dark matter candidates? 04:42:09 Raj Handique: Can we not directly get the nanoAOD root files and access it using ROOT and do analysis, completely skipping the need of CMSSW, as it is a flat ntuple? 04:43:50 vinny: Would great to see the raw sensor data to understand. Would be great if you could release a sample of raw data and the data in nanoAOD to understand it 04:45:28 Julie Hogan: @vinny thanks for the idea, we will think about whether we can do something like that after the workshop 04:50:02 Andrzej Konrad Siodmok: I have a question, there is a particle-flow Nano for JetHT. Lets say I know what jet algorytm was used in the analysis. I am wondering if i run fastjet on pfcandidates do I get the same jets as stored in the JetHT dataset or there is some information missing? 04:51:34 Julie Hogan: For those who are interested in a completion certificate (and anyone who finds a cool plot), please paste it in today's google doc! https://docs.google.com/document/d/1B_xaD3zbxERqOnA2DD1i6jUA9CEbhLhprYk5U1ZKuWk/edit 04:53:06 Andrzej Konrad Siodmok: Thanks! I would like to try it :) 04:53:39 Andrzej Konrad Siodmok: I would probably need some help ;) 04:55:20 Márcio Mateus Jr: When I run the comand to get the file name I get a list of files ending with thhe .root format, do I need all of them or just the first? 04:56:40 Annamalai S: Is it okay with a jupyter notebook to run the commands 04:57:15 Sahil Rawa: For those who are interested in a completion certificate (and anyone who finds a cool plot), please paste it in today's google doc! https://docs.google.com/document/d/1B_xaD3zbxERqOnA2DD1i6jUA9CEbhLhprYk5U1ZKuWk/edit didn't understand it 04:57:33 Ramakrishnan GK: How to get the infile? 04:58:26 Sahil Rawa: which plot i have to paste there ? 04:58:32 Julie Hogan: @Sahil Rawa the request is to make a screencapture of the plot that you are making right now 04:58:33 Márcio Mateus Jr: But all the files are the same? I mean, to get all data from the dataset, I need all those root files? 04:58:59 Julie Hogan: Ah, I misunderstood -- no, the files are all unique. If you want to get ALL the data you need ALL the files. 04:59:13 João de Felipe Andrade: My connection is kind of slow, so I couldn't test it with the bigger datasets, but I've noticed uproot.open takes one URI. Is there anything that handles the multi-file datasets transparently? 04:59:17 Julie Hogan: That is not the request for this activity, but in general for an analysis you want all the files. We will show this tomorrow 05:00:45 Matthew Bellis (he/him): https://uproot.readthedocs.io/en/latest/basic.html 05:00:58 Matthew Bellis (he/him): "Reading many files into big arrays" 05:01:28 Julie Hogan: @Ramakrishnan GK you can find a file name from the dataset you've chosen by clicking "Download" on the file indices at the bottom of the dataset record website. Then copy one of the file names you see 05:02:27 Daohan Wang: I have one question. It appears that the current NanoAOD and MiniAOD datasets only correspond to the 2015 and 2016 data. Will there be NanoAOD or MiniAOD data available for the 2018 and 2019 data in the future? If I want to use the 2018 dataset, are there any former workshops or resources available on using CMSSW? 05:05:25 Julie Hogan: On Thursday I'll post some slides with links to previous workshops that talk more about analyzing AOD or MiniAOD in CMSSW 05:07:28 Caue Evangelista De Sousa: Plots are done!