diff options
Diffstat (limited to 'docs')
| -rw-r--r-- | docs/notes.rst | 8 | ||||
| -rw-r--r-- | docs/process.rst | 25 |
2 files changed, 16 insertions, 17 deletions
diff --git a/docs/notes.rst b/docs/notes.rst index a06d4580..b8ab4a5d 100644 --- a/docs/notes.rst +++ b/docs/notes.rst @@ -5,8 +5,9 @@ Some notes about how to recompile packages from unstable: `<http://wiki.debian.org/DebianUnstable>`_ Openblas with Sandybridge target support is available in the experimental -repository in debian. I recompiled it as follows: -TARGET=SANDYBRIDGE USE_OPENMP=1 apt-get -b source openblas +repository in debian. I recompiled it as follows:: + + TARGET=SANDYBRIDGE USE_OPENMP=1 apt-get -b source openblas By default, it uses pthreads, instead of openmp, which was interfering with other use of openmp in my code. If we dont' specify the target, it @@ -27,7 +28,6 @@ Update serie after default Everything should be done right by running the script ``handle_default.py``. This is an example run after Sabine Oil -Default: +Default:: -:: python handle_default.py 103944 2015-06-23 15.875 100 diff --git a/docs/process.rst b/docs/process.rst index bc94226e..8feb6247 100644 --- a/docs/process.rst +++ b/docs/process.rst @@ -6,11 +6,11 @@ our directory ``//WDSENTINEL/share/CorpCDOs`` on Windows and ``/home/share/CorpCDOs`` on Linux.
Restarting the tasks server
---------------------------
+---------------------------
Most of the operation below rely on the task server being
operational. In the unlikely event that it stops, it can be restarted
-as follows: ::
+as follows::
ssh debian
cd /home/share/CorpCDOs/code/python
@@ -81,8 +81,8 @@ Running the scenarios in Intex * First make sure that Intex look for the sss files in the right
directory. Go into Tools, Options, Data Access and change it
- appropriately. (e.g. Z:\CorpCDOs\Scenarios\Intex
- curves_2013-07-22\sss)
+ appropriately. (e.g. ``Z:\CorpCDOs\Scenarios\Intex
+ curves_2013-07-22\sss``)
* In Intex Options, under Miscellaneous, Change the export settings to
"Tab delimited"
* Select the cusips of interest on the spreadsheet (make sure the
@@ -96,7 +96,7 @@ Running the scenarios in Intex * Then go onto the Cashflows tab and wait... Make sure format of Intex export
is in the format required--> ET: CLO_layout. Once it's done, export the cashflows into
``Scenarios/Prices_%Y-%m%-%d``.
-* Finally, on the server do: ::
+* Finally, on the server do::
cd /home/share/CorpCDOs/code/R
Rscript load_cf.R
@@ -111,7 +111,7 @@ Calibration The model calibration is triggered via a cronjob at 8am. If something
wrong happened or we forgot to save the data, it's possible to
-re-trigger it manually as follows: ::
+re-trigger it manually as follows::
cd /home/share/CorpCDOs/code/R
Rscript calibrate_tranches_MF.R %Y-%m-%d
@@ -126,39 +126,38 @@ scripts to bring the prices into the database. In windows - run R script ``\CorpCDOs\code\R\load_bloomberg_data.R``
-On server: ::
+On server::
cd /CorpCDOs/code/python/
python load_bloomberg_data.py
Install lossdistrib
-------------------
+-------------------
- If Lossdistrib package is out of date, install in server by: ::
+If Lossdistrib package is out of date, install in server by::
cd /home/share/CorpCDOs/code/R
R CMD INSTALL lossdistrib
Kick off daily tranche runs
---------------------------
-
::
+
cd /home/share/CorpCDOs/code/R
Rscript calibrate_tranches_BC.R -u -c runs.yml
Rscript tranches_RV_BC.R -u -c RVruns.yml
Start Juypter
-------------
-
::
+
cd /home/share/CorpCDOs/notebooks
myipython
To install RPostgres
---------------------
-To access serenitasdb, I need some R packages... These particular ones are from github
+To access serenitasdb, I need some R packages... These particular ones are from github ::
-::
devtools::install_github("rstats-db/DBI")
devtools::install_github("thrasibule/RPostgres@convert_dates")
|
