shians / cellbench Goto Github PK
View Code? Open in Web Editor NEWR package for benchmarking single cell analysis methods
License: GNU General Public License v3.0
R package for benchmarking single cell analysis methods
License: GNU General Public License v3.0
vignette("Introduction", package = "CellBench")
Warning message:
vignette ‘Introduction’ not found
Where can I find more detailed documentation, it does not seem to find the associated instance in the .rmd of the related ‘CellBench_data’ project.
At the moment task_error
objects are propagated for the purposes of recording errors in method calls. A class task_stop
could be implemented with similar functionality but based on custom stop conditions.
For example, task_stop
could be emitted based on a filtering step removing too many cells. Another use case would be if only the top 5/10 stepwise results based on some metric should be passed into the next method, a filtering function could replace all non-passing results with a task_stop
that contains some information for the reason for stop.
The key functionality is that task_stop
has the same functionality as task_error
in terms of propagation, in that it is done automatically and individual methods do not need to know how to handle them.
When applying a metric as a method, the results are a list and the metric type is in a column. The desire behaviour would be a vector column with the metric name and the values unlisted.
Check exported functions and
Check private functions and
Progress reporting would be useful for applying pipelines. Can do this at the method step checkpoints and/or BiocParallel level.
Either add column selection or exclude to pipeline_collapse
, most likely exclusion because final column set needs to retain order.
Once categories have been defined in #4, each category should have a shared conceptual framework to be explained in a vignette.
For categories of API defined in #4
Currently CellBench stores every result for each step before moving on to the next step. It's possible to reduce memory burden by running single pipelines to the final result before moving onto the next, reducing the amount of data that needs to be kept in memory.
The difficulty is in computing all required data only once and releasing them when they are no longer needed, this would follow a depth-first tree structure and it's not obvious how to implement it elegantly.
I think it's worth running goodpractice::goodpractice(".")
and BiocCheck::BiocCheck(".")
(I won't post the output here since it's rather long).
Some of the warnings are spurious/premature since the package is still in development (e.g. lack of unit tests) but others are worth fixing now (e.g., use seq_len()
or seq_along()
instead of 1:N
).
Overally, I take these more as indicators than rules. But you'll ultimately need to satisfy BiocCheck for submission to Bioconductor.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.