Link
Search
Menu
Expand
Document
DiFlow
Introduction
Functional Reactive Programming (FRP)
FRP for pipelines
home
NextFlow
Introduction
FRP in NextFlow
NextFlow DSL(2)
DiFlow
The NoPipeline approach
General Requirements and design principles
An abstract computation step
Toward implementation
Step by step
Step 1 - Operate on a *stream*
Step 2 - Operate on a stream in parallel
Step 3 - Operate on a stream using a `process`
Step 4 - How `map` is synchronous
Step 5 - Introduce an `ID`
Step 6 - Add a process parameter
Step 7 - Use a `Map` to store parameters
Step 8 - Use a `Map` with a process-key
Step 9 - Use a `ConfigMap` with a shell script
Step 10 - Running a *pipeline*
Step 11 - A more generic process
Step 12 - Map/reduce in NextFlow
Step 13 - Files as input/output
Step 14 - *Publishing* output
Step 15 - Make output files/paths unique
Step 16 - Where to put `params`?
Step 17 - Add the output file to `params`
Step 18 - Add the output filename to the triplet
Step 19 - Use a closure
Step 20 - The order of events in a stream
Step 21 - Is the *triplet* really necessary?
Step 22 - Toward generic processes
Step 23 - More than one input
Step 24 - `workflow` instead of `process`
Step 25 - Custom scripts
Step 26 - The missing link
Putting it all together
What is missing from DiFlow?
Parameter checks
Multiple output file references
Per-sample configuration
Appendix
Variables in `nextflow.config`
Reasons for an explicit *flow*
Resources
Default values
What is missing from DiFlow?
Parameter checks
Parameter checks
We currently do no checks on variables that are set through the
ConfigMap
.