Lightcurve recipes :)

[kashish2210]

lightcurve :: recpies

So here I have used the datasets from_here [NICER-HESRAC-CL.EVT] .

USING PLOTS [PLOTING EXAMPLES FOR LIGHTCURVE]

using Plots
events = readevents("ni1200120104_0mpu7_cl.evt", load_gti=true, sort=true) #here load_gti will be default true

output

Found GTI data: 16 intervals
GTI time range: 1.3253976595089495e8 to 1.3261337476368216e8
EventList with 21244574 times and energies

simple plot

plot(events)

output

Light curve with energy filtering

plot(events, energy_filter=(0,5e3))

output

Light curve with time range filtering

plot(events, tstart=1.32540e8, tstop=1.32546e8)

output

lightcurve with axis limits and show_gtis and show_btis

plot(events, 5, axis_limits=[1.32540e8, 1.32546e8, 5e3, 2e4], show_gtis=true, show_btis=true)

output:

plot(events, show_btis=true, bti_alpha=0.4,show_gtis=true, gti_alpha=0.3)

output:

gti1 = [1.32540e8 1.32580e8]
plot(events, gtis=gti1, show_gtis=true,gap_threshold=100)
#or u can use `plot(events, gti_file="gti.fits", show_gtis=true)`

output:

rebining

eventlist = readevents("ni1200120104_0mpu7_cl.evt")
lc = create_lightcurve(eventlist, 800)
plot(lc, 1000, show_original=true, original_alpha=1)

output :

with Poission errors

suggestion when looking for errors, first create a lightcurve

lc = create_lightcurve(events, 10)
plot(lc, show_errors=true,)#u can apply axis_lmit

output

Now with monal_testA :

with Gaussian errors

# Create light curve from events with smaller bins for smoother appearance
lc = create_lightcurve(events, 1)  # Much smaller bins (0.1 seconds instead of 1.0)
lc_times = lc.time
lc_counts = lc.counts

# Apply smoothing to get continuous-looking data
using StatsBase
# Option 1: Moving average smoothing
window_size = 1000
smoothed_counts = [mean(lc_counts[max(1, i-window_size):min(end, i+window_size)]) 
                   for i in 1:length(lc_counts)]

# Option 2: Alternative - use a simple running mean
# smoothed_counts = running_mean(lc_counts, window_size)

# Create more realistic errors based on smoothed data
σ = sqrt.(max.(smoothed_counts, 1))  # Poisson-like errors, avoid sqrt of 0

# Create confidence bands
upper_1σ = smoothed_counts .+ σ
lower_1σ = smoothed_counts .- σ
upper_2σ = smoothed_counts .+ 2 .* σ
lower_2σ = smoothed_counts .- 2 .* σ
upper_3σ = smoothed_counts .+ 3 .* σ
lower_3σ = smoothed_counts .- 3 .* σ

# Plot with confidence bands
p = plot(lc_times, smoothed_counts, 
         seriestype=:line, linewidth=2, color=:blue,
         label="Data", xlabel="Time (s)", ylabel="Counts",
         title="Gaussian Confidence Bands")

# Add confidence bands (from outer to inner)
plot!(lc_times, [upper_3σ lower_3σ], fillrange=[lower_3σ upper_3σ], 
      fillalpha=0.2, fillcolor=:gray, label="3σ", linewidth=0)
plot!(lc_times, [upper_2σ lower_2σ], fillrange=[lower_2σ upper_2σ], 
      fillalpha=0.3, fillcolor=:orange, label="2σ", linewidth=0)
plot!(lc_times, [upper_1σ lower_1σ], fillrange=[lower_1σ upper_1σ], 
      fillalpha=0.4, fillcolor=:lightblue, label="1σ", linewidth=0)

# Replot the main line on top
plot!(lc_times, smoothed_counts, seriestype=:line, linewidth=2, color=:blue, label="")

# Display the plot
display(p)

output

segmentation

using Plots
events = readevents("ni1200120104_0mpu7_cl.evt", load_gti=true, sort=true)
segments = create_segments(events, 10000.0)
plot(segments, 
     show_errors=false,
     show_segment_boundaries=true,
     segment_colors=[:blue, :red, :green, :orange, :purple])

output:

tagging @fjebaker @matteobachetti @stefanocovino

[kashish2210]

I have added a new plotting recipe gti_length_distribution

using Plots
bti=BTIAnalysisPlot(events)
plot(bti,bti_analysis=true)

output:

plot(BTIAnalysisPlot(events), bins=1400,min_length=1e-5,max_length=10000000)
plot(BTIAnalysisPlot(events), bins=10000)

output

Tagging @matteobachetti [this new recpies with max_length , min_length works nearly like using geom scale]