Tutorial Day 3 review

[martinberoiz]

Tutorial 3.1

• Step 13 (bilby.run.sampler()) spits out a couple of length "pink-boxes" that is a bit ominous. Maybe there is a verbosity switch in Bilby to disable this, or at the very least a bit of text telling the user this pink output is normal. DONE (see Make bilby silent #42)
• the 4th step says "tame theoretician", which is a bit demaining. I suggest saying "theoretician colleague". DONE (see Fix language issues and typos #37)
• Step 8 says "Draw 10000 samples from g(theta)" in the first comment, but the code actually draws 100,000 samples. Should be consistent, probably just change the comment. DONE (see Fix language issues and typos #37)
• Q1.1 the answer is still on the page, "Yes, these results are consistent with one another to within the standard deviation." DONE (see Remove partial answers in Tuto 3.1 #36)
• Q4.1 the answer is still on the page, "Both are close to 0.1, but there is some scatter." DONE (see Remove partial answers in Tuto 3.1 #36)
• Q3.1 the somewhat demeaning adjective is used again "tame theorist" ... I suggest changing this to "theorist colleague". DONE (see Fix language issues and typos #37)

Tutorial 3.2

• Bilby again produces the verbose pink boxes of output that would be good to disable or inform the user it is normal. DONE (see Make bilby silent #42)
• Step 4, the bilby version is printed out, but there is nothing in the comment saying what this should be (is this left-over debugging code?). For what it is worth, I got 2.4.0 on my local conda installed environment. WONT FIX (this is not really a problem)
• step 7 it says "As such, we can plot the data itself out:", should this say "As such, we can plot out the data itself:"? DONE (see Fix language issues and typos #37)
• step 13, the plot would benefit from a legend to distinguish the two curves. Same for step 15's plot of two curves

Tutorial 3.3

• Maybe in other tutorials as well as I didn't catch this until working on tutorial 3, the use of "confidence interval" is used, however the collaboration has a preference to use "credible interval" instead. DONE (see Fix language issues and typos #37)
• Step 3, the user is asked to download a file from the DCC. These files have additional documentation here about the special release of data for GWTC-1 that the user may wish to read.
• Step 9, 13, marginalized is misspelled, s -> z. This may be the case in other places but these are the first place I caught it. DONE (see Fix language issues and typos #37)
• Challenge question, it is noted that "The z-component of each component spin is defined as chi_1z = chi1 * cos(theta)_1. This is missing the mapping from chi1 to spin1 found in the samples file, and cos(theta)_1 to costilt1 in the samples file. Should document this so the student can find the correct samples to use.

[duboism]

Part of this were fixed in #36.

[duboism]

Part of this were fixed in #37.

[duboism]

There is a way to make bilby less verbose: bilby.core.utils.log.setup_logger(log_level='WARNING'). In this case, it will only print warnings and errors.

[duboism]

Part of this was addressed in #42.