User account for 6-BM-A using ipython
Install bluesky core packages first
As of 04/17/2019, python3.6 is the prefered version for running BlueSky.
conda install bluesky -c lightsource2-tag
then the apstools dependencies
conda install pyresttable -c prjemian
followed by installing apstools
conda install apstools -c aps-anl-dev
Before installing the metapackage jupyter
, it is recommended to pin the package tornado
to an older version until BlueSky dev team solve the related issue#1062.
To do so, create a file named pinned under the directory CONDA_INSTALL_DIR/env/ENVNAME/conda-meta
with the following content:
tornado<5
Then install jupyter
and matplotlib
with
conda install jupyter matplotlib
For update BlueSky packages, one can always do update with explicit channel name
conda update bluesky -c lightsource2-tag
Alternatively, a package configuration file .condarc
can be placed under HOME
(single env) or CONDA_INSTALL_DIR/envs/ENV_NAME
(multi-env) with the following content
channels:
- lightsource2-tag
- lightsource2-dev
- aps-anl-tag
- aps-anl-dev
- prjemian
- defaults
- conda-forge
NOTE: This is the recommended way to install BlueSky and associated dependencies.
Install bluesky and apstools with pip
pip install -U pip
pip install boltons mongoquery pims pyepics pyRestTable tzlocal jupyter suitcase matplotlib
pip install git+https://github.com/Nikea/historydict#egg=historydict \
git+https://github.com/NSLS-II/amostra#egg=amostra \
git+https://github.com/NSLS-II/bluesky#egg=bluesky \
git+https://github.com/NSLS-II/databroker#egg=databroker \
git+https://github.com/NSLS-II/doct#egg=doct \
git+https://github.com/NSLS-II/event-model#egg=event_model \
git+https://github.com/NSLS-II/ophyd#egg=ophyd \
git+https://github.com/NSLS-II/hklpy#egg=hklpy
pip install apstools
Similarly, the package tornado
need to be downgrade below 5.0 to avoid the runtime error.
Copy the configuration file configs/mongodb_config.yml
to HOME/.config/databroker/mongodb_config.yml
to enable meta-data handler backed by a MongoDB server.
The entry host need to be changed to the IP of the machine that hosts the MongoDB service.
The environment var IPYTHONDIR
needs to be set where the profile folder is, or simply create symbolic link of the deisred profile to ~/.ipython/
, which is the default location to store all IPython profiles.
>> export IPYTHONDIR=ipython_profiles/
Issue the following command in the terminal to run IPython with pre-configured environment for Tomo-characterization at 6-BM-A:
>> ipython --profile=s6bm