save_trajs support for fragmented data source

[clonker]

see markovmodel/pyemma_tutorials#198

[codecov]

Codecov Report

Merging #1514 (b2748ca) into devel (366ce03) will increase coverage by 0.00%.
The diff coverage is 96.80%.

@@           Coverage Diff           @@
##            devel    #1514   +/-   ##
=======================================
  Coverage   92.03%   92.04%           
=======================================
  Files         229      229           
  Lines       26804    26830   +26     
=======================================
+ Hits        24670    24695   +25     
- Misses       2134     2135    +1     

Impacted Files	Coverage Δ
pyemma/coordinates/clustering/tests/test_kmeans.py	98.00% <ø> (-0.01%)	⬇️
pyemma/coordinates/data/util/reader_utils.py	87.71% <66.66%> (-0.57%)	⬇️
pyemma/coordinates/api.py	88.11% <87.50%> (-0.11%)	⬇️
pyemma/coordinates/clustering/regspace.py	93.24% <92.85%> (+5.57%)	⬆️
pyemma/_base/serialization/tests/test_cli.py	100.00% <100.00%> (+3.33%)	⬆️
pyemma/coordinates/clustering/interface.py	89.25% <100.00%> (ø)
pyemma/coordinates/clustering/kmeans.py	94.42% <100.00%> (+0.09%)	⬆️
pyemma/coordinates/tests/test_save_trajs.py	98.87% <100.00%> (+0.19%)	⬆️
...emma/_ext/variational/solvers/tests/test_direct.py	88.88% <0.00%> (-8.89%)	⬇️
... and 4 more

---

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 366ce03...b2748ca. Read the comment docs.

[clonker]

The test failure is unrelated to this set of changes, following up with another PR

[thempel]

Can't this exception be caught explicitly?

[clonker]

with the next deeptime release yes :) right now it's set up a bit unfortunate with the rmsd metric so i can only catch it implicitly.