[WIP] Implement data cubes.

[matteobachetti]

This PR aims at producing data cubes out of SDTimager. The functionality is more or less there, but we need to change the code in a few points

• We need fast histogram functions for the increased data size.
• When we load new subscans, we have to preserve the spectral information.
• The histogram2d functions in imager.py:559-- have to be substituted by histogram3d functions, and the 2-d case has to work as a special case with a single bin along the 3rd dimension
• Calibration has to take into account this new format
• The output data format of images has to change slightly to accommodate these cubes, and revert to the 2d case for non-spectroscopic images.
• Desideratum: as the increased data size might fill up the memory, this might be a good moment to implement some sort of memory mapping.

[codecov-io]

Codecov Report

Merging #192 into master will not change coverage.
The diff coverage is 100%.

@@           Coverage Diff           @@
##           master     #192   +/-   ##
=======================================
  Coverage   93.77%   93.77%           
=======================================
  Files          23       23           
  Lines        5867     5867           
=======================================
  Hits         5502     5502           
  Misses        365      365

Impacted Files	Coverage Δ
srttools/utils.py	94.23% <100%> (ø)	⬆️
srttools/imager.py	94.05% <100%> (ø)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update a01059f...438c9f4. Read the comment docs.

[matteobachetti]

Current progress: most functionality works. Need to fix problems with the baseline subtraction (somehow, it is sometimes run on spectroscopic data instead of total power)