Today first public presentation of pylayers at CNRS GDR Isis meeting on UWB and localization

[Bernard Uguen]

Dear all,

Today was the first public presentation of the PyLayers platform @
Telecom Paris Tech.
It was a really high quality workshop. Many new ideas, many bright
presentations.

Not really easy to go back and forth between pdf and notebook.
What is particularly annoying is the requirement to evaluate each cell
sequentially.
If one miss the first import cell, got those disturbing errors which are
not real errors but
you begin to feel the demo effect malediction.

The proper workflow should be:

1) Main presentation on Desktop 1 - Notebook list on desktop 2
1) CTRL+2 Open the notebook
2) Evaluate all cells
3) Drag until the cell of interest where you need to make a real time
experimentation
to give an idea of the processing time. I think not much than one
evaluation per notebook.

It would be also interesting in the future to distinguish between :

+ Long notebooks for website documenation
+ Short notebooks for live presentation (4 at 5 cell not more) -
It is difficult to get the point and also it is hard
to navigate in this long notebook during the presentation.
Better to split in many
small notebooks

I think the talk was too dense. It could definitely be improved for the
next time. Including for example
more graphical information on available outputs.

My top priority now is to work on the antenna module.
I have fixed different bug during the short life time of my battery on
my train way back.

I realized that in fact there exist wo similar functions for reading trx
antenna data file from Satimo NFC

+ read_trx (deprecated) very bad implementation (I kept it because it is
an example how to use regular
expression for parsing ine of text)

+ readtrx this function parse the header and determine the following
information

Trx header structure

fmin fmax Nf phmin phmax Nphi thmin thmax Ntheta #EDelay
0 1 2 3 4 5 6 7 8 9
1 10 121 0 6.19 72 0 3.14 37 0

Not all the information from the header are read (this is wrong) . Most
of the information is retrieved from the data. This requires an assert
to check
the integrity of the header with the data.

In the future it will be good to replace this header by an ini file

Best regards,

Bernard

--
Bernard Uguen
Professeur des Universités
Equipe Propagation Localisation
Université de Rennes 1
IETR UMR 6164 (Bat 11 D - Pièce 128)
Avenue du Général Leclerc, Campus de Beaulieu
35042 - Rennes Cedex
tél. : 02 23 23 60 33 - fax : 02 23 23 56 16