The Strip Instrument Database

Introduction

Every simulator of a scientific experiment needs an «instrument database», i.e., a way to access the specification of the instrument to be simulated. The kind of information stored in this database is the following:

How many detectors are present in the instrument, and what are their characteristics in terms of noise and placement on the focal plane;
What is the response of the optical elements, i.e., the so-called beam function;
Which frequencies can be measured by each detector (the so-called bandpass);
Ect.

Stripeline implements an instrument database using YAML files and two data structures, Detector and Horn, plus some functions to easily access the information in them.

The following example initializes an object of type InstrumentDB with the values referred to the standard STRIP instrument:


julia> db = InstrumentDB()
InstrumentDB(55 horns, 66 detectors)

This command loads the YAML files provided in the Stripeline repository and initializes the db object. As db is a struct, its field can be accessed with the usual dot notation. The two fields in db are focalplane and detectors. They are both dictionaries, associating horn names to Horn objects and detectors IDs to Detector objects, respectively:

julia> db.focalplane["I0"]
Horn I0, module 0 (indigo), STRIP31:
    Orientation: [0.0000, 0.0000, 1.0000]
    FWHM (X/Y): 0.3470 × 0.3440 °
    Spillover: 0.016704 (main), 0.013964 (sub)
    Cross-polarization: 44.53 dB
    Directivity: 54.76 dBi
    Ellipticity: 1.0080

julia> db.detectors[2]
Detector STRIP02 (Q band):
    Center frequency: 43.39 ± 0.13 GHz
    Bandwidth: 7.58 ± 0.87 GHz
    Noise temperature: 31.5 ± 3.92 K
    Knee frequency: 90.0 ± 40.0 mHz (Q), 70.0 ± 40.0 mHz (U)

A number of high-level functions ease the access of the fields in a InstrumentDB object:

detector returns a Detector structure, containing the details of a polarimeter;
bandpass returns a BandshapeInfo structure, containing the shape of the bandpass of a detector;
spectrum returns a SpectrumInfo
fknee_hz returns the knee frequency of the 1/f noise for the I, Q, and U signals, adapted to the brightness temperature of the load being observed by the detector;
tnoise returns the noise temperature for the I, Q, and U components.

The structure Detector uses three structures to organize its data in a hierarchical way:

BandshapeInfo
SpectrumInfo
NoiseTemperatureInfo

All these structures know how to show themselves on the REPL:

julia> db.detectors[2].bandshape
Bandshape:
    Center frequency: 43.39 ± 0.13 GHz
    Bandwidth: 7.58 ± 0.87 GHz
    Frequency range: [38.00, 50.00] GHz (121 points)
    Test ID: [355, 356]
    Analysis ID: 89

julia> db.detectors[2].spectrum
Noise spectrum:
    Slope: I = 0.8585 ± 0.0000, Q = 0.6420 ± 0.0030, U = 0.3680 ± 0.0030
    Knee frequency: I = 0.0 ± 0.0 mHz, Q = 90.0 ± 40.0 mHz, U = 70.0 ± 40.0 mHz
    White noise: Q = 5.0 ± 1.0 mK^2 Hz, U = 2.6 ± 0.6 mK^2 Hz
    System brightness temperature: 20.0 K
    Test ID: 357
    Analysis ID: 15

julia> db.detectors[2].tnoise
Noise temperature:
    Tnoise: 31.5 ± 3.9 K
    Estimates: [26.7, 27.1, 27.6, 28.2, 30.1, 31.1, 32.0, 33.6, 34.0, 35.1, 36.2, 39.7] K
    Test IDs: []
    Analysis IDs: [105, 106]

Some of them have the ability to be plotted using Plots.jl:

julia> using Plots

julia> plot(db.detectors[2].bandshape)
Plot{Plots.GRBackend() n=1}

For more information about the fields in the structures listed above, as well as their meaning, keep reading.

Structures

Stripeline.InstrumentDB — Type

STRIP instrument database

The "database" contains information about feed horns and polarimeters:

The field focalplane is a dictionary (mapping) associating the string identifying a horn (e.g., I0) with a Horn structure;
The field detectors is a dictionary associating the ID of the polarimeter (e.g., 2 stands for STRIP02) with a Detector structure.

You should usually create an object of this kind using the default constructor, which parses a set of YAML files containing the real parameters of the instrument.

Examples

julia> db = InstrumentDB();

julia> print("Number of horns in the database: $(length(keys(db.focalplane)))")
Number of horns in the database: 55

julia> print("Number of polarimeters in the database: $(length(keys(db.detectors)))")
Number of polarimeters in the database: 66

Visualization

You can produce a table describing the contents of the instrument database using show and passing text/markdown as MIME type:

db = InstrumentDB()
show(stdout, MIME("text/markdown"), db)

The table can be converted to other formats (HTML, LaTeX, Microsoft Word, …) using commonly-available tools, e.g., Pandoc.

Field	Type	Meaning
`name`	String	Name of the horn, e.g., `I0`
`id`	Int	Unique number of the horn, starting from 1
`polid`	Int	Unique ID of the polarimeter associated with the horn
`polarizerid`	Int	Unique ID of the polarizer+OMT associated with the horn
`moduleid`	Int	Number of the horn within the module, from 0 to 6
`color`	String	Name of the color associated with the module
`orientation`	Array{Float64}	3D vector containing the orientation of the horn in the sky
`fwhm_x_deg`	Float64	FWHM of the beam along the X axis, in degrees
`fwhm_y_deg`	Float64	FWHM of the beam along the Y axis, in degrees
`main_spillover`	Float64	Main reflector spillover
`sub_spillover`	Float64	Sub-reflector spillover
`xpd_db`	Float64	Cross-polarization, in dB
`directivity_dbi`	Float64	Directivity, in dBi
`ellipticity`	Float64	Ellipticity

Field	Type	Meaning
`id`	Int	Integer ID of the polarimeter, e.g., `2` for `STRIP02`
`name`	String	Full name of the polarimeter, e.g., `STRIP02`
`band`	String	Band: it can either be `Q` or `W`
`bandshape`	BandshapeInfo	Information about the bandpass response
`spectrum`	SpectrumInfo	Information about the noise spectrum (white noise and 1/f noise)
`tnoise`	NoiseTemperatureInfo	Information about the noise temperature

Field	Type	Meaning
`center_frequency_hz`	Float64	Estimate for the center frequency, in Hz
`center_frequency_err_hz`	Float64	Estimated error on the center frequency, in Hz
`bandwidth_hz`	Float64	Estimated bandwidth, in Hz
`bandwidth_err_hz`	Float64	Estimated error on the bandwidth, in Hz
`lowest_frequency_hz`	Float64	Lowest frequency of the bandshape in `response`, in Hz
`highest_frequency_hz`	Float64	Highest frequency of the bandshape in `response`, in Hz
`num_of_frequencies`	Int	Number of samples in `response`
`bandshape`	Array{Float64,1}	Profile of the bandshape (pure numbers)
`bandshape_error`	Array{Float64,1}	Estimated error on the profile of the bandshape
`test_id`	Array{Int,1}	ID of the unit-level test used to characterize the bandshape
`analysis_id`	Int	ID of the unit-level analysis used to characterize the bandshape

Field	Type	Meaning
`slope_i`	Float64	The slope ($\alpha$) of the 1/f component of the noise in the I signal
`slope_i_err`	Float64	Error associated with the value of `slope_i`
`slope_q`	Float64	Same as `slope_i`, but for the Q signal
`slope_q_err`	Float64	Error associated with the value of `slope_q`
`slope_u`	Float64	Same as `slope_i`, but for the U signal
`slope_u_err`	Float64	Error associated with the value of `slope_u`
`fknee_i_hz`	Float64	Knee frequency of the I signal, in Hz
`fknee_i_err_hz`	Float64	Error associated with the value of `fknee_i_hz`
`fknee_q_hz`	Float64	Knee frequency of the Q signal, in Hz
`fknee_q_err_hz`	Float64	Error associated with the value of `fknee_q_hz`
`fknee_u_hz`	Float64	Knee frequency of the U signal, in Hz
`fknee_u_err_hz`	Float64	Error associated with the value of `fknee_u_hz`
`wn_i_k2_hz`	Float64	White noise level for the I signal, in K^2 Hz
`wn_i_err_k2_hz`	Float64	Error associated with the value of `wn_i_k2_hz`
`wn_q_k2_hz`	Float64	White noise level for the Q signal, in K^2 Hz
`wn_q_err_k2_hz`	Float64	Error associated with the value of `wn_q_k2_hz`
`wn_u_k2_hz`	Float64	White noise level for the U signal, in K^2 Hz
`wn_u_err_k2_hz`	Float64	Error associated with the value of `wn_u_k2_hz`
`i_fit_parameters_k2_hz`	Vector{Float32}	Fit coefficients for the I spectrum in K²/Hz, or `nothing`
`q_fit_parameters_k2_hz`	Vector{Float32}	Fit coefficients for the Q spectrum in K²/Hz, or `nothing`
`u_fit_parameters_k2_hz`	Vector{Float32}	Fit coefficients for the U spectrum in K²/Hz, or `nothing`
`pwr_cov_matrix_k2`	Symmetric{4}	Covariance matrix of the signals Q1, Q2, U1, U2 (PWR) in K² or `nothing`
`dem_cov_matrix_k2`	Symmetric{4}	Covariance matrix of the signals Q1, Q2, U1, U2 (DEM) in K² or `nothing`
`iqu_cov_matrix_k2`	Symmetric{3}	Covariance matrix of I = ∑PWR / 4, Q = (Q1 + Q2) / 2, U = (U1 + U2) / 2
`load_temperature_k`	Float64	System brightness temperature used during the tests (in K)
`test_id`	Int	ID of the unit-level test used to characterize the bandshape
`analysis_id`	Int	ID of the unit-level analysis used to characterize the bandshape

Field	Type	Meaning
`tnoise_k`	Float64	Noise temperature computed from `tnoise_values_k`, in K
`tnoise_err_k`	Float64	Error associated with `tnoise_k`, computed from `tnoise_values_k`
`test_ids`	Array{Int,1}	List of unit-level test IDs used to estimate the noise temperature
`analysis_ids`	Array{Int,1}	List of unit-level analysis report IDs used to estimate the noise temperature
`values_k`	Array{Float64,1}	List of noise temperatures estimated from the tests

The Strip Instrument Database

Introduction

Structures

High-level access functions

Loading custom databases