reduced packages.tsv to only python, r and the essential r-packages. …

…Everything else is now inside csp_wiesner_johannes repo

README.md

@@ -6,7 +6,7 @@ Using `.yml` files as recipes to create conda environments is already a good ste
 
 ## Presolved environments using Github Actions
 
-This repository contains a subfolder called `environments` with a `packages.tsv` file in it. The `packages.tsv` file contains all packages that the [**Complex Systems in Psychiatry Lab**](https://www.zi-mannheim.de/en/research/departments-research-groups-institutes/psychiatry-psychotherapy/researchgroups-psychiatry-e/complex-systems-in-psychiatry.html) needs for its research (but can be easily adapted to your own needs). `environments` again contains two subfolders `windows-latest` and `ubuntu-latest`. These folders contain pre-solved files of the package specifications found in `packages.tsv` for both Windows and Ubuntu by running a Github Actions workflow that uses tcy & Micromamba behind the scenes to create these files.
+This repository contains a subfolder called `environments` with a `packages.tsv` file in it. The `packages.tsv` file by default only contains Python, R and the essential R-packages (but can be easily adapted to your own needs, see next section). `environments` also contains two subfolders `windows-latest` and `ubuntu-latest`. These folders contain files that have the solved package specifications found in `packages.tsv` for both Windows and Ubuntu by running a Github Actions workflow that uses tcy & Micromamba behind the scenes to create these files.
 
 Users can download this repository and use the  `ubuntu-latest_solved.yml` or `windows-latest_solved.yml` files to create a conda environment.
 

environments/packages.tsv

@@ -1,66 +1,4 @@
 package_name	version	installation_command	package_manager	conda_channel	area	link	necessity	description	comment	language	bug_flag
-nimare		pip install nimare	pip		statistical modelling	https://nimare.readthedocs.io/en/stable/	required	Used to derive statistical images from neurosynth.org		Python	
-python	<3.10	conda install -c conda-forge python	conda	conda-forge	language	https://anaconda.org/conda-forge/python	required	the python programming language itself		Python	
-datalad		conda install -c conda-forge datalad	conda	conda-forge	data wrangling	https://www.datalad.org/	required	Allows you to download and manage datalad datasets	https://www.datalad.org/	Python	
-tableone		conda install -c conda-forge tableone	conda	conda-forge	data wrangling	https://github.com/tompollard/tableone	required	Package for plotting descriptive statistics		Python	
-plotly-orca		conda install -c plotly plotly-orca	conda	plotly	dependency	https://plotly.com/python/orca-management/	required	Optional addon  for plotly to export static images	Plotly people now recommend to use kaleido instead (https://plotly.com/python/static-image-export/). So this should be changed to kaleido	Python	
-bokeh		conda install -c bokeh bokeh	conda	bokeh	dependency	https://docs.bokeh.org/en/latest/	required	Optional addon  for holoviews to create interactive plots		Python	
-spyder		conda install -c anaconda spyder	conda	conda-forge	ide	https://www.spyder-ide.org/	required	scientific python IDE		Python	
-scikit-learn		conda install -c conda-forge scikit-learn	conda	conda-forge	machine learning	https://scikit-learn.org/stable/index.html	required	machine learning libary	similar to nilearn, this can be installed both using conda and pip (sklearn developers don't recommend one over the other), note that sciki-learn is a dependeny from nilearn and so will automatically be installed when installing nilearn	Python	
-nilearn		conda install -c conda-forge nilearn	conda	conda-forge	neuroimaging	https://nilearn.github.io/index.html#	required	Statistics for NeuroImaging in python	there seems to be a discussion to use conda instead of pip among the nilearn developers but currently pip seems to be recommended	Python	
-nipype		conda install -c conda-forge nipype	conda	conda-forge	neuroimaging	https://nipype.readthedocs.io/en/latest/index.html	required	python interface package for common neuroimaging software		Python	
-plotly		conda install -c plotly plotly	conda	plotly	plotting	https://plotly.com/python/	required	Plotting library similar to seaborn but allows for interactive plots	We also install plotly-orca, so we are able to plot static images (.jpg,.png, etc.)	Python	
-holoviews		conda install -c pyviz holoviews	conda	pyviz	plotting	https://holoviews.org/index.html	required	wrapper around common python plotting libraries (note that installation command will also install bokeh and pyviz, which is the recommended way from holoviews)	Johannes niplot module needs this	Python	
-statannotations		conda install -c conda-forge statannotations	conda	conda-forge	plotting	https://github.com/trevismd/statannotations	required	A library that allows you to plot figures using seaborn and also computing significance tests and adding asterisks	Note, that the original package statannot is not maintained anymore, this is why we use statannotations which is a maintained branch	Python	
-fastcluster		conda install -c conda-forge fastcluster	conda	conda-forge	statistical modelling	https://anaconda.org/conda-forge/fastcluster	required	This package can speed up hierarchical clustering methods such as used in seaborn.clustermap		Python	
-pingouin		conda install -c conda-forge pingouin	conda	conda-forge	statistical modelling	https://pingouin-stats.org/index.html#	required	Implements commong statistical tests (ANOVA, t-test, etc.)		Python	
-pymer4		conda install -c conda-forge pymer4	conda	ejolly	statistical modelling	https://eshinjolly.com/pymer4/index.html	required	A library that allows you to compute mixed models (with the lme4 R package) using python (and the ryp2 package)	pymer4 depends on R (and lme4) and on rpy2 so this will automatically be installed in the environment (see: https://eshinjolly.com/pymer4/installation.html)	Python	
-network_control		pip install network_control	pip		statistical modelling	https://github.com/BassettLab/control_package	required	Package to compute network control measures		Python	
-cca-zoo		pip install cca-zoo	pip		statistical modelling	https://cca-zoo.readthedocs.io/en/latest/index.html	required	implementing canonical correlation analysis in python		Python	
-bctpy		pip install btcpy	pip		statistical modelling	https://github.com/aestrivex/bctpy	required	A library that allows you to compute all sorts of topological measures on connectivity matrices		Python	
-gower		pip install gower	pip		statistical modelling	https://pypi.org/project/gower/	required	A package that allows you to compute Gower Distance Matrices		Python	
-antropy		pip install antropy	pip		statistical modelling	https://github.com/raphaelvallat/antropy	optional	AntroPy is a Python 3 package providing several time-efficient algorithms for computing the complexity of time-series.		Python	
-mne		conda install -c conda-forge mne	conda	conda-forge	statistical modelling	https://mne.tools/stable/index.html	optional	Python package for analyzing MEG/EEG data		Python	
-pyvis		pip install pyvis	pip		plotting	https://pyvis.readthedocs.io/en/latest/index.html	optional	Visualizing graphs/networks in an interactive fashion	This will automatically install networkx as a dependency	Python	
-distro		pip install distro	pip		system	https://pypi.org/project/distro/	required	distro provides information about the OS distribution it runs on, such as a reliable machine-readable ID, or version information.	This package is useful to detect on which specific distro your current python script is running. For example you might want to run the script both on your personal machine and on the server. In this case paths might change and you want to have an option to automatically change paths by detecting the OS	Python	
-abagen		pip install abagen	pip		statistical modelling	https://abagen.readthedocs.io/en/stable/installation.html	required	abagen lets you get information about gene receptor densities using the Allen Human Brain Atlas and a custom atlas of your choice		Python	
-oct2py		conda install -c conda-forge oct2py	conda	conda-forge	system	https://pypi.org/project/oct2py/	required	Oct2py allows use to run Octave from Python (so you can execute .m files)		Python	
-hvplot		conda install -c pyviz hvplot	conda	pyviz	plotting	https://hvplot.holoviz.org/index.html	required	hvplot allows to use all the pandas plotting functions (e.g. df.hist()) but outputs interactive instead of static figures		Python	
-groupyr		pip install groupyr	pip		statistical modelling	https://richiehalford.org/groupyr/index.html	required	Allows you to compute group sparse regression / classification		Python	cross-platform
-clustergrammer		pip install clustergrammer	pip		plotting	https://clustergrammer.readthedocs.io/clustergrammer_py.html	required	Let’s you create interactive clustermaps		Python	
-nisupply		pip install nisupply	pip		data wrangling	https://github.com/JohannesWiesner/nisupply	required	Johannes Wiesner’s repository to deal with unstructured neuroimaging datasets		Python	
-pybids		conda install -c conda-forge pybids	conda	conda-forge	data wrangling	https://bids-standard.github.io/pybids/	required	Pybids helps to query BIDS-datasets		Python	
+python		conda install -c conda-forge python	conda	conda-forge	language	https://anaconda.org/conda-forge/python	required	the python programming language itself		Python	
 r-essentials		conda install -c conda-forge r-essentials	conda	conda-forge	data wrangling	https://tinyurl.com/54j96j5x	required	The R Essentials bundle contains approximately 200 of the most popular R packages for data science, including the IRKernel, dplyr, shiny, ggplot2, tidyr, caret, and nnet. It is used as an example in the following guides.		R	
-r-repr		conda install -c conda-forge r-repr	conda	conda-forge	dependency	https://anaconda.org/conda-forge/r-repr	required	This is required if you want to use the Variable Inspector from Jupyter-Lab to be able to inspect R-objects	https://github.com/lckr/jupyterlab-variableInspector#requirements-for-r-functionality	R	
-rtools		conda install -c r rtools	conda	r	dependency	https://anaconda.org/r/rtools	required		Why do I added this? Check why exactly this has to be installed. Current hypothesis: This is needed for devtools.This package is necessary to be able to install R-packages from GitHub. This package is not available in Linux	R	linux
-lckr-jupyterlab-variableinspector		pip install lckr-jupyterlab-variableinspector	pip		dependency	https://github.com/lckr/jupyterlab-variableInspector	required	Adds a variable inspector to jupyterlab	Optional addon for jupyterlab that adds a variable explorer. I couldn't install it on Windows. See issues from JohannesWiesner on GitHub (you can filter for my name in the repo on GitHub). But on Linux it works fine	R	windows
-r-devtools		conda install -c conda-forge r-devtools	conda	conda-forge	development	https://anaconda.org/conda-forge/r-devtools	required	Installing R-packages from GitHub	This package is necessary to be able to install R-packages from GitHub (aka. that can not be installed using conda or CRAN). After installed you can run commands like devtools::install_github()	R	
-rstudio		conda install -c r rstudio	conda	r	ide	https://www.rstudio.com/	required	An IDE for R	Currently not recommended to install R-studio in a conda environment because this will downgrade your R-version to 3.x.x. The current alternative is jupyter-lab. This should be checked regularily to see if a new update was released. Maybe in the future it wil become in a more recent version in conda-forge?	R	cross-platform
-jupyterlab		conda install -c conda-forge jupyterlab	conda	conda-forge	ide	https://jupyter.org/	required	An alternative IDE to RStudio, which might be faster and offers a more pythonic experience		R	
 r-base		conda install -c conda-forge r-base	conda	conda-forge	language	https://anaconda.org/conda-forge/r-base	required	the R programming language itself		R	
-r-reticulate		conda install -c conda-forge r-reticulate	conda	conda-forge	data wrangling	https://cran.r-project.org/web/packages/RcppCNPy/vignettes/UsingReticulate.pdf	required	reticulate allows you to read in data generated from Python (e.g. .npy files)		R	
-scikit-image		conda install -c conda-forge scikit-image	conda	conda-forge	statistical modelling	https://scikit-image.org/	required	We need that for otsu thresholding		Python	
-pyarrow		conda install -c conda-forge pyarrow	conda	conda-forge	data wrangling	https://arrow.apache.org/docs/python/install.html	optional	pyarrow is required to save pandas data frames as .feather files (which is the format to exchange data between python and R)	Not recommended to install that, because the 'R-side' of this version can not be installed on the CIMH servers (not sure if this issue still exists). Besides, one can also use rpy2 to exchange files.	Python	cross-platform
-xmltodict		pip install xmltodict	pip		data wrangling	https://pypi.org/project/xmltodict/	optional	working with xml files	this is required for Johannes pycat library	Python	
-yellowbrick		conda install -c districtdatalabs yellowbrick	conda	districtdatalabs	plotting	https://www.scikit-yb.org/en/latest/index.html	optional	Builds upon seaborn and has some advanced plotting classes such as Radviz (https://www.scikit-yb.org/en/latest/api/features/radviz.html)		Python	
-dash		conda install -c conda-forge dash	conda	conda-forge	plotting	https://dash.plotly.com/introduction	optional	Used for nct_tutorial from Johannes		Python	
-dash-bootstrap-components		conda install -c conda-forge dash-bootstrap-components	conda	conda-forge	plotting	https://dash-bootstrap-components.opensource.faculty.ai/docs/quickstart/	optional	Used for nct_tutorial from Johannes	Allows to create dash-apps using a grid-layout (e.g. rows and columns)	Python	
-adjustText		pip install adjustText	pip		plotting	https://github.com/JohannesWiesner/pycat/blob/master/pycat.py	optional	adjust text in matplotlib figures		Python	
-dash-cytoscape		pip install dash-cytoscape	pip		plotting	https://dash.plotly.com/cytoscape	optional	Used for nct_tutorial from Johannes	Allows to plot networks in dash	Python	
-brainrender		pip install brainrender	pip		plotting	https://github.com/brainglobe/brainrender	optional	plotting 3D brains		Python	
-semopy		pip install semopy	pip		statistical modelling	https://semopy.com/	optional	Used for structural equation modelling in python	Bug: scikit-learn has now to be installed as scikit-learn, not sklearn (https://gitlab.com/georgy.m/semopy/-/issues/35), once issue is solved we can re-include semopy again	Python	
-gemmr		pip install gemmr	pip		statistical modelling	https://github.com/murraylab/gemmr	optional	A library that allows you to use Witten et als. PMA function via rpy2		Python	
-resample		pip install resample	pip		statistical modelling	https://github.com/scikit-hep/resample	optional	Allows you to do permutation and bootstrapping resampling	This requires a C-compiler to be installed on your machine. On CIMH machines this can be tedious to install	Python	cross-platform
-pytorch		conda install -c pytorch pytorch	conda	pytorch	machine learning	https://anaconda.org/pytorch/pytorch	optional	Tensor library for deep learning using GPUs and CPUs		Python	
-tensorboardx		conda install -c conda-forge tensorboardx	conda	conda-forge	machine learning	https://anaconda.org/conda-forge/tensorboardx	optional	Allows researchers to use simple interface to log events within PyTorch		Python	
-neurokit2		conda install -c conda-forge neurokit2	conda	conda-forge	statistical modelling	https://github.com/neuropsychology/NeuroKit	optional	NeuroKit2 is a user-friendly package providing easy access to advanced biosignal processing routines	This package is specifically needed to compute multiscale sample entropy (there are also other packages though, that can do this or might include this in the future which are also smaller: https://github.com/nikdon/pyEntropy, https://github.com/ELIFE-ASU/PyInform)	Python	
-fmriprep-docker		pip install fmriprep-docker	pip		neuroimaging	https://fmriprep.org/en/20.2.0/installation.html#the-fmriprep-docker-wrapper	optional	This package makes the invocation of fmriprep easier		Python	
-jupyter-book		conda install -c conda-forge jupyter-book	conda	conda-forge	ide	https://jupyterbook.org/en/stable/intro.html	optional	Used for creating workshop webpages. Not needed for any data analysis		Python	
-sklearn-pandas		conda install -c conda-forge sklearn-pandas	conda	conda-forge	data wrangling	https://github.com/scikit-learn-contrib/sklearn-pandas	optional	Nötig, um scikit-learn transfomer direkt auf pandas data frames anzuwenden		Python	
-arrow		conda install -c conda-forge --strict-channel-priority r-arrow	conda	conda-forge	data wrangling	https://arrow.apache.org/docs/r/	optional		Note: This currently seems to work for Windows but struggles with installation on Linux! Arrow is required to save data frames as .feather files (which is the format to exchange data between python and R). This documented error happens on the server: https://arrow.apache.org/docs/r/articles/install.html#package-failed-to-build-c-dependencies	R	cross-platform
-r-matrixstats		conda install -c conda-forge r-matrixstats	conda	conda-forge	data wrangling	https://anaconda.org/conda-forge/r-matrixstats	optional		required for reproducing Xia et al. CCA analysis	R	
-r-pma		conda install -c conda-forge r-pma	conda	conda-forge	statistical modelling	https://anaconda.org/conda-forge/r-pma	optional	Package to conduct sparse canonical correlation analysis	Why do I added this? This is the same as install.packages('PMA') that can also be found in this spreadsheet? Choose either this package or install PMA package with CRAN, but not both. If you can succesfully use the gemmr package and if you can rerun Xia et al. script, then you don't have to use install.package('PMA'')	R	
-PMA		install.packages('PMA')	cran		statistical modelling	https://cran.r-project.org/web/packages/PMA/index.html	optional		required for reproducing Xia et al. CCA analysis	R	
-lme4		install.packages('lme4')	cran		statistical modelling	https://cran.r-project.org/web/packages/lme4/lme4.pdf	optional	Fitting mixed models in R	Note, that when using the python package  pymer4, you will already have lme4 installed, so this installation would become obsolete. You might also need to install stamod (https://community.rstudio.com/t/error-message-when-trying-to-install-lmer-package-and-load-lmer-library/85418)	R	
-r-dendextend		conda install -c conda-forge r-dendextend	conda	conda-forge	plotting	https://r-charts.com/part-whole/circular-dendrogram/	optional	Allows to customize dendrograms (e.g. coloring branches, adding labels)	Should be installed together with circlize	R	
-r-circlize		conda install -c conda-forge r-circlize	conda	conda-forge	plotting	https://r-charts.com/part-whole/circular-dendrogram/	optional	Allows to draw circular dendrograms	Should be installed together with dendextend	R