Decode

This release fixed the following bugs:

1. [creation of default scantype](https://github.com/deshima-dev/decode/commit/953bd5ac749803cd4033f3b85e5c4c3a4fd2d84c)
2. [invalid function name](https://github.com/deshima-dev/decode/commit/f37ef88197679bcac56e05a3d3fb2a6fac0e0495)

array function
dc.concat
+ [**Bug Fixed**] `xref` and `yref` of the concatenated array is adjusted to the 1st array.

i/o function
dc.io.loaddfits
+ A new keyword of `shuttle` is implemented for shuttle observations. When this is True, x coordinates of off-point and on-point are specified via keywords of `xmin_off`, `xmax_off`, `xmin_on`, and `xmax_on`.

model funtion
+ **Calculation speed gets much faster!**
dc.models.chopper_calibration
+ A new calibration method has been implemented. This can be used in the same way as dc.models.rsky_calibration.
python
dc.models.chopper_calibration(onarray, offarray, rarray, Tamb, mode='mean')

plot function
dc.plot.plotpsd
+ PSD plot.
python
plotpsd(data, dt, ndivide=1, window=hanning, overlap_half=False, ax=None, **kwargs)


dc.plot.allanvar
+ Allan variance plot.
python
plotallanvar(data, dt, tmax=10, ax=None, **kwargs)


cube function
dc.makecontinuum
+ Values of channel coordinates of the continuum (`masterid`, `kidid`, `kidtp`, and `kidfq`) have been changed.

dc.savefits
+ [**Bug Fixed**] The header for the az/el data has been corrected.

i/o function
dc.io.loaddfits
+ A kwargs of `findR` has been implemented. When this is True, loaddfits will automatically find R positions according to some user-defined parameters. `Rth` is the threshold above which data will be categorized as R, and `skyth` is the threshold below which data will be categorized as SKY. This calculation depends on one representative channel, which specified via a keyword of `ch`. `cutnum` determines the number of data points which will be not used at the edges.
+ A kwargs of `still` has been implemented. When this is True, loaddfits will change scantypes to SCAN and OFF. The cycle is specified via a keyword of `period`.

model function
dc.models.pca
+ [**Bug Fixed**] [**IMPORTANT**] The implementation of PCA had problems, but the bugs are removed.
+ None can be used as `mode`.

dc.models.r_subtraction
+ Apply R-subtraction.
python
dc.models.r_subtraction(onarray, offarray, rarray, mode='mean')


dc.models.r_division
+ Apply R-subtraction.
python
dc.models.r_division(onarray, offarray, rarray, mode='mean')


utils
dc.utils.psd
+ Calculate PSD.
python
dc.utils.psd(data, dt, ndivide=1, window=hanning, overlap_half=False)


dc.utils.allan_variance
+ Calculate Allan variance.
python
allan_variance(data, dt, tmax=10)

plot function
dc.plotspectrum
- [**Backward Compatibility Lost**] A keyword of `peak` for `ytick` is replaced to `max`.
- You can specify `mean` for `ytick` now.

cube function
dc.tocube
- You can specify the unit of arguments via a keyword of `unit`. The default is `deg`.
python
arcsecond
dc.tocube(array, gx=10, gy=10, unit='arcsec')
arcminute
dc.tocube(array, gx=0.1, gy=0.1, unit='arcmin')

- The default value of `xc` and `yc` is changed to the value of `xref` and `yref` in the cube.
- [**Bug Fixed**] After converted from array, `coordsys`, `xref`, and `yref` in the cube are correctly inherited.

dc.makecontinuum
- [**Backward Compatibility Lost**] Previously, this function has returned a xarray whose shape is x/y. For now, the returned xarray has the shape of x/y/ch, but the length of ch is only 1.

i/o function
dc.io.savefits
- Some header information is added to output FITS files, but some of them have just default values.
- When the length of ch is 1, i.e. for the continuum data, you can choose whether the resulting dimensions are x/y or x/y/ch via a keyword of `dropdeg`. This is the same implementation as CASA.
- **Resulting FITS file can be loaded in CASA for now.**

cube function
dc.tocube
+ [**Bug Fixed**] Direction of right ascention in 3D cubes is corrected.
+ Noise map information is included to cubes. It can be accessed by `cube.noise`. When you want a noise map, you just need to run `dc.io.savefits(cube.noise, fitsname)`.

plot function
dc.plot.timestream
+ you can select the type of x axis via a keyword of `xtick`.
python
real time
dc.plot.plottimestream(array, ax, xtick='time', **kwargs)
time index
dc.plot.plottimestream(array, ax, xtick='index', **kwargs)


i/o function
dc.io.loaddfits
+ you have more choices for `loadtype`, such as `amplitude`, `phase`, and `linphase`.