notebook.jl

### A Pluto.jl notebook ###
# v0.19.26

using Markdown
using InteractiveUtils

# ╔═╡ 524225b9-d165-47ab-9c49-dffbd93393aa
# hideall
begin
	using Pkg
	Pkg.activate(@__DIR__)
    Pkg.instantiate()
	using CairoMakie, Revise, LHC_AGC, JSON3, FHist, UnROOT
end

# ╔═╡ 97e9c1f4-d444-4db6-9037-28a68f2deccd
using BenchmarkTools, PlutoUI # hide

# ╔═╡ 5f403020-c4ba-4814-926d-389e2347e9d2
# by default download 1 file per sample
LHC_AGC.download_data() # this might take a long time with the current

# ╔═╡ 23afac95-4797-4b5c-b7f1-6340313925d9
tt_sample = first(readdir(LHC_AGC.BASE_PATH[]; join=true))

# ╔═╡ c82d8346-fa92-4159-b4ac-85ecfe035297
md"""
## Introduction: exploring $t\bar{t}$
"""

# ╔═╡ c0175af3-2903-4def-8599-da934fc28803
const tt_tree = LazyTree(tt_sample, "Events"); # hide

# ╔═╡ 87017bd9-8622-43c5-aa44-f24a6c1b39b1
njet_hist = Hist1D((length(x) for x in tt_tree.Jet_pt), 0:16; overflow=true)

# ╔═╡ 72be7e4c-d85a-4795-ba36-4e83771ba160
# chop off the overflow bin
integral(restrict(njet_hist, 0, 15))

# ╔═╡ 92cd3a46-10bf-4b52-ad1a-ae07459830c8
with_terminal() do
	# this is a lazy generator, evident since it has 0 allocation
	global _c = @time (count(>(25), x) for x in tt_tree.Jet_pt)
end

# ╔═╡ 33b8830a-8365-4a37-ad12-180113101700
with_terminal() do
	# and the loop over this generator is fully inferred, which means it's fast
	@code_warntype Hist1D(_c, 2:10)
end

# ╔═╡ dbfd0d3c-2b83-48ba-b2df-ba93a4a7d34d
Hist1D(_c, 2:10)

# ╔═╡ f80e56a0-2ffb-44c8-a11d-6922c96d3188
md"""
## $t\bar{t}$ kinematic reconstruction of the top mass.

You can find the full function in the `main_looper.jl` file, and look below this cell to look at the result of calling the function. Here we present a snippt view of the main three parts of the main looper:

### Basic cuts
```julia
	# single lepton requirement
	# Avoid allocation:
	# - bad example: `length(electron_pt .> 25)`
	if count(>(25), electron_pt) + count(>(25), muon_pt) != 1
		continue
	end

	# at least 4 jets
	(; jet_pt) = evt
	# can't avoid allocation since we need this mask later
	jet_pt_mask = jet_pt .> 25
	count(jet_pt_mask) < 4 && continue

	# at least 2 btag
	# use view to avoid allocation
	jet_btag = @view evt.jet_btag[jet_pt_mask]
	count(>=(0.5), jet_btag) < 2 && continue
	(; jet_px, jet_py, jet_pz, jet_mass) = evt
```
"""

# ╔═╡ e484f0c5-73fc-40a3-b875-ee8a9e939e17
# let's not weight this one
res = @time LHC_AGC.get_histo(:ttbar; wgt=1.0, n_files_max_per_sample=1);

# ╔═╡ 516e76f1-cc21-49b5-8a02-be47318e78ad
begin
	nominal_hist = res["nominal"]["4j2b"]
	stairs(nominal_hist,
		axis=(xlabel="mass (GeV)", ylabel="# events", 
		limits=(0,375,0,30000),
		xticks=0:50:400,
		yticks=0:5000:80000,
		)
	)
	errorbars!(nominal_hist)
	vlines!(175, color=:grey, linestyle=:dash)
    text!(180, 20, text = L"m_t = 175 \mathrm{GeV}")
	current_figure()
end

# ╔═╡ Cell order:
# ╠═524225b9-d165-47ab-9c49-dffbd93393aa
# ╠═97e9c1f4-d444-4db6-9037-28a68f2deccd
# ╠═5f403020-c4ba-4814-926d-389e2347e9d2
# ╠═23afac95-4797-4b5c-b7f1-6340313925d9
# ╟─c82d8346-fa92-4159-b4ac-85ecfe035297
# ╠═c0175af3-2903-4def-8599-da934fc28803
# ╠═87017bd9-8622-43c5-aa44-f24a6c1b39b1
# ╠═72be7e4c-d85a-4795-ba36-4e83771ba160
# ╠═92cd3a46-10bf-4b52-ad1a-ae07459830c8
# ╠═33b8830a-8365-4a37-ad12-180113101700
# ╠═dbfd0d3c-2b83-48ba-b2df-ba93a4a7d34d
# ╟─f80e56a0-2ffb-44c8-a11d-6922c96d3188
# ╠═e484f0c5-73fc-40a3-b875-ee8a9e939e17
# ╠═516e76f1-cc21-49b5-8a02-be47318e78ad