Metadata-Version: 2.1
Name: tdr
Version: 0.0.1
Summary: TriCCS Data Reduction
Home-page: https://bitbucket.org/jin_beniyama/triccs-data-reduction/src/master/
Author: Jin BENIYAMA
Author-email: beniyama@ioa.s.u-tokyo.ac.jp
Maintainer: Jin BENIYAMA
Maintainer-email: beniyama@ioa.s.u-tokyo.ac.jp
License: UNKNOWN
Platform: UNKNOWN
Description-Content-Type: text/markdown

# TriCCS Data Reduction (TDR)
[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)

[developer mail](<mailto:beniyama@ioa.s.u-tokyo.ac.jp>)


## Overview
Data reduction of data taken with Seimei/TriCCS could be done in this repository.  
Though optimized for Seimei/TriCCS, 
you can apply it for imaging data taken with other high-speed cameras
(Tomo-e Gozen etc.).

### Procedures
1. Calibration (dark subtraction, flat-field correction)


2. Stacking fits by mean, median, etc.


3. Splitting fits
(only for photometry using `Moving Object Photometry (movphot)` ([bitbucket](https://bitbucket.org/jin_beniyama/movphot/src/master/)) )

4. Common ID search

## Installing
```
# Install from PyPI
pip install tdr
```
or
```
# Install with `setup.py`
git clone https://jin_beniyama@bitbucket.org/jin_beniyama/triccs-data-reduction.git
python setup.py install
```


## Usage
Here g-band data taken with Seimei/TriCCS is considered.
All band data could be analyzed in the same way.

Fits data taken with TriCCS have a format like `TRCS00005180.fits`.

First 4 characters `TRCS` means the instrument *TriCCS*,
the next 7 characters are the exposure ID 

and the last 1 character is band identical number 
(`0` for g-band, `1` for r-band and `2` for i/z-band).

After each reduction stage, a prefix is added to the filename.
History can be checked in fits header as well.


### 1. Calibration
Here, consider the situation:
exposure time for an object frame is 10 s,
for a flat frame is 1 s,
and for dark frames are 1s and 10s (all in g-band) 
like below.

1. object `TRCS00000010.fits` (10 s)
2. dark for flat `TRCS00000020.fits` (1s)
3. dark for object `TRCS00000030.fits` (10s)
4. flat `TRCS00000040.fits` (1s)

Dark subtraction and flat-field correction are done as follows.


First, create master dark frame, which has prefix `d` like `dTRCS00000020.fits`.

The maximum count frame is not used for stacking,
which leads to avoiding cosmic rays or fast-moving object contamination.
```
[usage]
# Create master dark
makedark (3-d dark)

[example]
# Create master dark for flat
makedark TRCS00000020.fits
# Create master dark for object
makedark TRCS00000030.fits
```

Next, create a master normalized flat frame using master dark for a flat frame.
, which has prefix `f` like `fTRCS00000040.fits`.

The maximum count frame is not used for stacking as well.

```
[usage]
# Create master flat
makeflat --flat (3-d flat) --dark (2-d master dark)

[example]
# Create master flat
makeflat --flat TRCS00000040.fits --dark dTRCS00000020.fits
```

Finally, reduce an object frame using both master dark and flat frames. 

The reduced object frame has prefix `r` like `rTRCS00000010.fits`.

```
[usage]
# Do dark subtraction and flat-field correction
reduce --obj (3-d object) --dark (2-d master dark) --flat (2-d master flat)

[example]
# Do dark subtraction and flat-field correction
reduce --obj TRCS00000010.fits --dark dTRCS00000020.fits --flat fTRCS00000040.fits
```

### 2. Stacking
Output fits have format like 
`maxrTRCS00000010.fits` (max),
`minrTRCS00000010.fits` (min),
`meanrTRCS00000010.fits` (mean) and
`medianrTRCS00000010.fits` (median).

```
[usage]
# Maximum stacking
stackfits (3-d reduced fits) max
# Minimum stacking
stackfits (3-d reduced fits) min
# Mean stacking
stackfits (3-d reduced fits) mean
# Median stacking
stackfits (3-d reduced fits) median

[example]
# Maximum stacking
stackfits rTRCS00000010.fits  max
# Minimum stacking
stackfits rTRCS00000010.fits  min
# Mean stacking
stackfits rTRCS00000010.fits  mean
# Median stacking
stackfits rTRCS00000010.fits  median
```

### 3. Splitting [in preparation]
If you are going to use `Moving Object Photometry (movphot)`[(bitbucket)](https://bitbucket.org/jin_beniyama/movphot/src/master/) for photometry,
3-d fits cube should be split into multiple 2-d fits.

```
[usage]
# Mask pixels and split fits into multiple 2-d fits
mask_split (3-d fits)

[example]
# Mask pixels and split fits to multiple 2-d fits
mask_split rTRCS00000010.fits
```
Output fits are as follows (when the number of frames is 3).
The masked and splitted frames hav suffix `ms` like `rTRCS00000010ms0001.fits`.


### 4. Common ID search
If the wcs pasting failed for some fits,
it is necessary to extract common ID fits.


```
[example]
# Extract common fits ID from g,r,i bands list (glist.txt, rlist.txt, ilist.txt)
cat glist.txt | awk '{print substr($0,17,3)}' > gID.txt
cat rlist.txt | awk '{print substr($0,17,3)}' > rID.txt
cat ilist.txt | awk '{print substr($0,17,3)}' > iID.txt
# Create common ID list 
commonIDsearch gID.txt rID.txt iID.txt --pre rTRCS00001260ce0 
--post w.fits > glist_common.txt
```

## Dependencies
This library is depending on `NumPy`.
Scripts are developed on `Python 3.7.10` and `NumPy 1.19.2`.


## LICENCE
This software is released under the MIT License.



