ply-compatible float64 types

[Tofull]

Description

.ply files were not written with "double" coordinates even though the xyz coordinates columns were in float64 format.
I was dealing with ply files in geospatial context and required to save a .ply file. I hit this issue, so here is the PR.

This seems to have been a pain for the community:

• As far as I remember, float64 caused the export to binary formats (i.e ply) to fail.
  float type conversion #146 (comment)

• The problem I found was that even though I could export/import PLY files with float64 values inside pyntcloud, those files weren't correctly loaded in other programs such as Three.js or CloudCompare so I just decided to force the float32 conversion because any of the logic in pyntcloud was affected by that.
  float type conversion #146 (comment)

• Can I save a pointcloud as double precision  #234

Proposed changes

• make float64 compatible with ply files (written ply files with float64 (double) coordinates can be opened by cloudcompare)

• make property_format mapping better (written ply files can now exploit all the scalar data types a .ply file can support)
  

• try downcasting column if column dtype is not compatible with PLY specification

• fix most of flake8 issues (1 remaining)

• backward compatible with Add support for bool type in PLY files #309

References :

• numpy built-in scalar types : https://numpy.org/devdocs/reference/arrays.scalars.html#built-in-scalar-types
• ply format description : http://paulbourke.net/dataformats/ply/

[Tofull]

Was wondering if the downcast propagation should be done on a copy of the points dataframe instead of the points dataframe itself.
If true, we should add points = points.copy() at the beginning of pyntcloud.io.ply.write_ply function.