ifcimglib

Install

Run:

pip install .

to install this package.

How to use

cif2lmdb can be used as follows:

Usage: cif2lmdb [OPTIONS] CIF

Options:
  --output FILE       Output filename. If not set, cif-filename is taken with
                      lmdb extension.

  --channels INTEGER  Images from these channels will be extracted. Default is
                      to extract all. 1-based index.

  --names TEXT        Names to assign to channels.
  --debug             Show debugging information. Limits output to 100 first
                      cells.

  --overwrite         Overwrite lmdb if it exists.
  --targets-npy FILE  Numpy binary file containing targets.
  --skip-npy FILE     Numpy binary file containing instances to be skipped.
  --help              Show this message and exit.

Here is an example command:

cif2lmdb --channels 1 --channels 6 --channels 9 --names BF --names SSC --names BF2 --output tmp.lmdb --debug --overwrite input.cif

It takes input.cif as input and outputs output.lmdb, an lmdb-file containing 100 (see debug flag) 3-channel images with names BF, SSC and BF2.

Please see the imglmdb.ipynb notebook for usage examples of the imglmdb package.

Docker

Docker images are available for cif2lmdb (tag cif2lmdb), the jupyter lab environment (tag jupyter-lab-env), and the notebooks in this repository (tag notebooks).

For using cif2lmdb, run:

docker run --rm -v /path/to/data/dir:/data maximlippeveld/ifcimglib:cif2lmdb [OPTIONS] /data/example.cif

For using the jupyter environment, run:

docker run --it -v /path/to/data/dir:/data -v /path/to/your/code/dir:/app -p [your-port]:8888 maximlippeveld/ifcimglib:jupyter-lab-env

Fur using the notebooks in this repository, run:

docker run --it -v /path/to/data/dir:/data -p [your-port]:8888 maximlippeveld/ifcimglib:notebooks