DONT DELETE ITEMS, USE ITEMS LISTED AS FIXED FOR NEW FEATURE LIST

after everything, make sure new data goes through fine


If there are multiple flats close to a observation, it wont choose the correct one. I.e. if there is a 67 order and a 68 order it may grab 67 which crashes.

FIXED way to edit specific observations if the pipeline is failing for it somehow. 
create your own analyzed data object and push it then the pipeline will skip.

bad shift reasoning:  output/30495/58762/9957401_report.pdf

verify open license

function to calculate mean of star's shk for adjusting to MT wilson

epi eri is bad test star, reversal ruins windowed shifting sometimes

Check if larger windows in multi function means larger integration window

FIXED Lab spec in intermediate plot is broken somehow

NOT DOING THIS B/C NRES GIVES IN NM-swap code and output to angstroms

FIXED-remove idl remenants and other obsolete code/variables

DONESKI once folder order set, place functions on top for easy finding

cofirm with TBrown or Ricky that we should always use filename for star over the star name from fits header

FIXED Convolutions take the most time(outside of plotting)

FIXED-Easy to use plot function to take this pipeline data and print it with other data

!!!!!!!!!!!!!!!!!!!!!!!Try catch around single observations as well as stars to allow for large data sets to be processed despite potential errors.

FIXED Old spectra format from NRES may not be running both observations. THIS IS DUE TO TWO OBS PER FOLDER 

I BELIEVE FINE check calc_shk is using windows function properly, may be accessing array wrong.

MOSTLY FIXED, TEST NEEDED set up optional radial velocity shifting(if statement to ignore cross-corr align)
DOCUMENT THIS

SEMI FIXED store a stars bad data file and skip it next re-run
currently just printed but should go into a file which is always read from, with debug option to skip bad

FIXED change all array shapes to python style to remove transpose

Question In making the flat we use the lamda range of 380-420 but it seems like we are using significantly less.

FIXED - better documentation for alpha value and

MAYBE FIXED - changed mk_flatolap bounds for each order. This decides how much of each order the flat takes and adds.
It used to have different values for each order but they were hacked in. Now I used roughly the lowest and highest
from the original lists for all orders. Seems the have normal flats.
BAD FLATS THAT COULD BE CAUSED BY THIS CHANGE - 
30495/58705/3358909
30495/58705/3435641


make pipeline callable on remote server

create function to load data that comes from np.savez NRES_SHK_Pipeline.py


FIXED?!??!?!******window alignment functionality*****

Investigate and tweek bad detection settings. Find simple star with knwon good and bad data to decide on specific values for auto-correlation.

License update