|
From: lemoene <le...@bi...> - 2011-11-17 10:49:28
|
Hi Alfonso
Stop me when this gets too much (or we can go off-list - but in
public here we might get some useful feedback - are you there Jean;-)
Am not sure about your team's available capacity, but if you are going
to dive into the Bika 3 code any time soon, I recommend tackling any of
these semi-detached outstanding items first:
1. Pdf writing
In Inkosi LIMS 2 pdfs are only used as results publication
preference - client contacts may select it from amongst all the
other publication preferences and receive their analysis results as
pdf attachments to emails. Eventually, LIMS 3.1 say, we would also
make it possible to save and email all reports as pdf.
We used ReportLab <http://www.reportlab.com/software/opensource/>
for Inkosi and would consider it in conjunction with pisa
<http://pypi.python.org/pypi/pisa/> this time round for its
Plone/Zope integration and forward looking approach.
2. Graphing
In Inkosi, we use Zope's RenderableCharts
<http://www.contentmanagementsoftware.info/zope/RenderableCharts>
and Reportlab again, as well as Gary Strangman's stats.py for the
stats calculations required for QC distribution curves etc.
R <http://www.r-project.org/> is the most obvious improvement to
stats.py. For their ease of integration with Plone, consider:
Matplotlib http://matplotlib.sourceforge.net/
http://www.scipy.org/Wiki/Cookbook/Matplotlib/Matplotlib_and_Zope
PyChart http://home.gna.org/pychart/
http://plone.org/documentation/kb/add-charts
Graphs used in Inkosi/Health 0.3:
Analysis results over time per sample point/patient
Reference results per analysis over time
Reference results distribution
% Relative error for duplicates
See QC reports and graphs
<http://www.bikalabs.com/helpcentre/manual/bikalims20/qc/qcreports>
re the last 3 in the Inkosi manual.
3. Working time calendar
Written up in the Inkosi calendar here: Bika Calendar Manual
<http://www.bikalabs.com/helpcentre/manual/bika-calendar-manual>. It
is currently only used to determine accurate turn around times,
excluding non-working hours, weekends and holidays, but has to be
approached carefully so as not to exclude future scheduling
requirements:
Scheduling sampling in Bika Water, some of it recurring - say
sampling the same river on a weekly basis. The calendar will in
this case be used to simply trigger the automated creation of
Analysis request, Sampling rounds and instructions, and alerting
Samplers to it.
Most complicated and notoriously challenging: to predict
accurate expected turnaround times for clients based on the
loads of instruments and analysts. Bika 3.5?-)
See you on the hub!-)
Bests
l
|