Data Analysis¶

The micro tomography computing infrastructure for the Imaging group is located at 2-BM:

Station	Name	Product	Part list	Model	Quote
2-BM	tomo 1-2	MNJ15421064	cluster	Supermicro 740GP-TNRT node	cluster_quote
2-BM	tomo 3	MNJ15421064	cluster_01	Supermicro 740GP-TNRT node	cluster_quote_01
2-BM	tomo 4-5	MNJ16235754	cluster_02	Supermicro 220GQ-TNAR+ node	cluster_quote_02
2-BM	tomodata1	MNJ15508749	disk_array_01	SYS-220U-TNR Storage	disk_array_quote_01
2-BM	tomodata2	MNJ18897861	disk_array_02	SYS-220U-TNR Storage	disk_array_quote_02

At the APS¶

Your raw data are automatically copied from the detector to the analysis computer (handyn in this example) under the folder /local/data/YYYY-MM/PI_lastName.

Manual¶

To manually reconstruct a data set, use the tomopy cli tool.

[tomo@tomo1,~]$ bash
[tomo@tomo1,~]$ conda activate tomopy

then for help:

[tomo@tomo1,~]$ tomopy recon -h

To do a test reconstruction type:

[tomo@tomo1,~]$ tomopy recon --file-name /local/data/YYYY-MM/PI_lastName/file.h5

Automatic¶

To setup a reconstruction to start and publish automatically the results on a google slide with tomolog, edit tomorec_log

[tomo@tomo1,~]$ bash
[tomo@tomo1,~]$ conda activate tomocupy
(tomocupy) [tomo@tomo1,~]$ sublime ~/bin/tomorec_log

by updating the –presentation-url to match the new google slide url

#!/usr/bin/bash
tomocupy recon --file-name $1 --remove-stripe-method fw --reconstruction-type full --rotation-axis-auto auto --find-center-end-row 1500
tomolog run --presentation-url https://docs.google.com/presentation/d/1YuxMttfW8w2sfwbaw634R3_LgPIsaHblz4Lrsjzn6ufQ/edit?usp=sharing --file-name $1 --beamline 2-bm --zoom [1,2,4]

then type:

(tomocupy) [tomo@tomo1,~]$ auto_rec /local/data/YYYY-MM/PI_lastName/

any new raw dataset uploade in /local/data/YYYY-MM/PI_lastName/ will be automatically reconstructed and results will be published on a google slide using tomolog.

At your home institution¶

Install the following:

Download and install anaconda python for your operative system.
Create a conda environment:

$ conda create -n tomopy python=3.9

Activate the newly created conda environment:

$ conda activate tomopy

Install tomopy:

$ conda install --channel conda-forge tomopy

Install dxchange:

$ conda install -c conda-forge dxchange

Install tomopy cli:

$ git clone https://github.com/tomography/tomopy-cli.git
$ cd tomopy-cli
$ python setup.py install

For Windows installation of tomopy-cli watch this video.

Install tomopy cli dependecy:

pip install opencv-python

To run a reconstuction you can now run:

$ tomopy recon --file-name /data/file.h5

Mosaic¶

For samples larger than the field of view we collect multiple data sets consisiting of overlapping tiles to form a mosaic. To reconstruct these type of data please use tile command-line-interface for mosaic tomography data processing.