Characteristics of the acm bias
Last updated:Jul17,2020

The acm bias frames are usually taken during the afternoon HET Opeatrions checkout.

 
 # Remember BaseDir file   
[sco@vdas 2018-2020]$ cat BaseDir
/hetdata/data

I typically run this analysis on vdas.
[sco@vdas 2018-2020]$ pwd  
/home/mcs/sco/ACM_BIAS_STUDY_Jul2020/2018-2020

To get the list of dates I want to review: 
% datelist.sh 2018 -ny 3 > list.2018-2020
Here I have made a list for 3 years of acm bias images (2018 to 2020) 

To run the initial compilation of bais data: 
[sco@vdas 2018-2020]$ multi_acm_bias list.2018-2020

This job took about 4h45m to complete and collected bias statistics from 495 nights. A summary of the data comse in the form of a table file: multi_acm_bias.parlab,table. I made some initial plots (below) to illustaret some of the results. Note that for for these plots with xyplotter I experimented with using a local version of the file named matplotlibrc. Here are some notes from that:

 

See my notes in /home/sco/Installs/install_sco2019_20200305/codes/python/plotting/Notes_from_SCO/README.jul12_2020 

Later I found how I change rcParams values within the code (from my interactive cursore code).
#=========================================================
# experimental
paramso = {'savefig.format'       : 'png',
          'figure.figsize'       : [9.0, 9.0],
          'font.size'            : 15,
          'axes.formatter.limits': [-3, 3]}
plt.rcParams.update(paramso)
#=========================================================

So I stuck this in pxy_SM_plot.py:
#=========================================================
# experimental ----- Jul2020
paramso = {'savefig.format'       : 'png',
          'figure.figsize'       : [12.0, 9.0]}
plt.rcParams.update(paramso)
#=========================================================

Now I get a much better size for plots that go into webpages (and that I can see!).

The mean bias values derived from acm bias frames covering a period of about three years.

Here we plot the same mean bias level shown in the previous figure, but on the X-axis we plot the UT time (in floating point hours) during the night.

Individual images over many nights.

I have modified the multi_acm_bias so that the "-s" optional flag will invoke acm_bias_singles that will break the table file into individual images.

 
I do this work in: 
/home/sco/BIAS_Study_Jul2020/Runs_vdas_Jul17_2020/Test1

I have 4 basic sets broken done by year:
2017(Sep-Dec)       cyan 
2018(Jan-May)       red 
2019(Jan-Dec)       green 
2020(Jan-Jul)       blue 

Here are the columns that are in  acm_bias_singles
col  name     Description                                            Good range 
1    bmean    Mean BIAS signal (adu)                                 1350  1430 
2    bme      Mean error about mean BIAS signal (adu)
3    bsig     Standard deviation about mean BIAS signal (adu)
4    npix     Number of pixels for BIAS determination
5    d2000    Days since Jan1,2000                                   6400  7600 
6    djan1    Days since Jan1
7    uth      UT in hours
8    temp     Ambient Temperature (deg C)                             -5    35 
9    year     Year
10   mon      Month number
11   day      Day number (in month)
12   image    acm image basename

I summarize some of the plots made with this data in a discussion of the mean acm bias levels over time.

Back to calling page