Skip to content

Commit

Permalink
Merge pull request #107 from DaGeibl/dev
Browse files Browse the repository at this point in the history 
renamed variables in lq_cut, changed expression handling
  • Loading branch information
@fhagemann
fhagemann authored Feb 21, 2025
2 parents 113fc5f + 295b064 commit 9e1f713
Show file tree
Hide file tree
Showing 4 changed files with 427 additions and 224 deletions.
145 changes: 128 additions & 17 deletions ext/LegendSpecFitsRecipesBaseExt.jl
View file
Original file line number Diff line number Diff line change
Expand Up @@ -1305,16 +1305,16 @@ end

# recipe for the lq_drift_time_correction report

@recipe function f(report::NamedTuple{(:lq_prehist, :lq_report, :drift_prehist, :drift_report, :lq_box, :drift_time_func, :DEP_left, :DEP_right)}, e_cal, dt_eff, lq_e_corr, plot_type::Symbol)
@recipe function f(report::NamedTuple{(:lq_report, :drift_report, :lq_box, :drift_time_func, :dep_left, :dep_right)}, e_cal, dt_eff, lq_e_corr, plot_type::Symbol)

# Extract data from the report
DEP_left = report.DEP_left
DEP_right = report.DEP_right
dep_left = report.dep_left
dep_right = report.dep_right
box = report.lq_box

#cut data to DEP
dt_DEP = dt_eff[DEP_left .< e_cal .< DEP_right]
lq_DEP = lq_e_corr[DEP_left .< e_cal .< DEP_right]
dt_dep = dt_eff[dep_left .< e_cal .< dep_right]
lq_dep = lq_e_corr[dep_left .< e_cal .< dep_right]

# Plot configuration: 2D histogram
xlabel := "Drift Time"
Expand All @@ -1330,7 +1330,7 @@ end


if plot_type == :DEP
# Create 2D histogram with filtered data based on DEP_left and DEP_right
# Create 2D histogram with filtered data based on dep_left and dep_right

# dynamic bin size dependant on fit constraint box
t_diff = box.t_upper - box.t_lower
Expand All @@ -1345,19 +1345,21 @@ end

@series begin
seriestype := :histogram2d
dt_DEP, lq_DEP
dt_dep, lq_dep
end
elseif plot_type == :whole
# Create 2D histogram with all data
colorbar_scale := :log10

# dynamic bin size
xmin = ustrip.(quantile(dt_eff, 0.001))
xmax = ustrip.(quantile(dt_eff, 0.999))
xstep = (xmax - xmin) / 200
ymin = ustrip.(quantile(lq_e_corr[.!isnan.(lq_e_corr)], 0.005)) #filters NaNs for quantile
ymax = ustrip.(quantile(lq_e_corr[.!isnan.(lq_e_corr)], 0.94)) #filters NaNs for quantile
ystep = (ymax - ymin) / 200
# dynamic bin size dependant on fit constraint box
t_diff = box.t_upper - box.t_lower
lq_diff = box.lq_upper - box.lq_lower
xmin = box.t_lower - 1*t_diff
xmax = box.t_upper + 1*t_diff
xstep = (xmax - xmin) / 400
ymin = box.lq_lower - 8*lq_diff
ymax = box.lq_upper + 8*lq_diff
ystep = (ymax - ymin) / 400
nbins := (xmin:xstep:xmax, ymin:ystep:ymax)

@series begin
Expand Down Expand Up @@ -1392,7 +1394,8 @@ end
# Evaluate drift_time_func over the full range of drift time (x-axis)
dt_range = range(xmin, xmax, length=100) # 100 points across the x-axis
lq_fit = report.drift_time_func.(dt_range) # Apply the linear function to the full dt range

xlims := (xmin, xmax)
ylims := (ymin, ymax)
dt_range, lq_fit

end
Expand Down Expand Up @@ -1420,7 +1423,11 @@ end
# 2D histogram for LQ Cut
xlabel := "Energy"
ylabel := "LQ (A.U.)"
nbins := (0:6:3000, -2.0:0.012:4.0)
#lq bins dependant on cut value
ymin = -5 * cut_value
ymax = 10 * cut_value
ystep = (ymax - ymin) / 500
nbins := (0:6:3000, ymin:ystep:ymax)
colorbar_scale := :log10
color := :viridis
legend := :bottomright
Expand Down Expand Up @@ -1476,8 +1483,112 @@ end

@series begin
label := "Cut Fraction"
0:3:3000, h_diff
0:3:3*length(h_diff)-1, h_diff
end

elseif plot_type == :fit
# Fit plot
xlabel := ""
ylabel := "Counts"

ylabel := "Normalized Counts"
margins := (4, :mm)
framestyle := :box
legend := :topleft
xlims = (ustrip(Measurements.value(report.fit_report.μ - 5*report.fit_report.σ)), ustrip(Measurements.value(report.fit_report.μ + 5*report.fit_report.σ)))
xlims := xlims
@series begin
label := "Data"
subplot --> 1
report.fit_report.h
end
@series begin
color := :red
subplot --> 1
label := "Normal Fit (μ = $(round_wo_units(report.fit_report.μ, digits=2)), \n σ = $(round_wo_units(report.fit_report.σ, digits=2)))"
lw := 3
bottom_margin --> (-4, :mm)
ustrip(Measurements.value(report.fit_report.μ - 5*report.fit_report.σ)):ustrip(Measurements.value(report.fit_report.σ / 1000)):ustrip(Measurements.value(report.fit_report.μ + 5*report.fit_report.σ)), t -> report.fit_report.f_fit(t)
end

@series begin
seriestype := :vline
label := "Cut Value"
subplot := 1
linewidth := 2
color := :red
[cut_value]
end

if !isempty(report.fit_report.gof)
link --> :x
layout --> @layout([a{0.7h}; b{0.3h}])
@series begin
seriestype := :hline
ribbon := 3
subplot --> 2
fillalpha := 0.5
label := ""
fillcolor := :lightgrey
linecolor := :darkgrey
[0.0]
end
@series begin
seriestype := :hline
ribbon := 1
subplot --> 2
fillalpha := 0.5
label := ""
fillcolor := :grey
linecolor := :darkgrey
[0.0]
end
@series begin
seriestype := :scatter
subplot --> 2
label := ""
title := ""
markercolor --> :black
ylabel := "Residuals (σ)"
xlabel := "LQ (A.U.)"
link --> :x
top_margin --> (-4, :mm)
ylims := (-5, 5)
xlims := xlims
yscale --> :identity
yticks := ([-3, 0, 3])

bin_midpoints = report.fit_report.gof.bin_centers
residuals = report.fit_report.gof.residuals_norm
bin_midpoints, residuals
end
end

elseif plot_type == :sideband
# Sideband histograms
xlabel := "Lq (A.U.)"
ylabel := "Counts"

@series begin
seriestype := :stepbins
label := "Peak"
report.temp_hists.hist_dep
end

@series begin
seriestype := :stepbins
label := "Sideband 1"
report.temp_hists.hist_sb1
end

@series begin
seriestype := :stepbins
label := "Sideband 2"
xlims := quantile(filter(isfinite, lq_class), 0.05), quantile(filter(isfinite, lq_class), 0.95)
report.temp_hists.hist_sb2
end


end
end

Expand Down
Loading

0 comments on commit 9e1f713

Please sign in to comment.