FsLab is a curated set of templates, libraries and scripts-as-journals support for doing data science with F#.
See https://fslab.org for getting started. One way is as follows:
dotnet new -i FsLab.Templates
dotnet new fslab-journal -lang F# -n Experiment1
then
cd Experiment1
build html
build show
build run (watches for changes)
edit Experiment1.fsx or open Experiment1.fsproj
The project produces three things:
FsLab.fsx
load script for a consistent set of packages, plus some extra formattersThe source files in the repository are organized as follows:
Directory or file | Comment |
---|---|
misc |
Icons and other non-source-code things |
src/fsx2html |
Source for the DLL in the fsx2html NuGet package |
src/Html |
HTML formatters for literate scripts not included in other packages |
src/Text |
Text formatters for literate scripts not included in other packages |
src/Shared |
Styling and server support for literate scripts not included in other packages |
src/experiments |
Item templates |
src/journal |
Journal project template |
src/dotnet-templates |
Extra files for 'dotnet new' template support |
src/vs-templates |
Extra files for Visual Studio template support |
src/FsLab.fsx |
Script included in the FsLab NuGet package |
src/*.nuspec |
NuGet files for building the packages |
build.fsx |
FAKE script that does all the magic (below) |
If you want to be able to build FsLab Journal template, you'll need Visual Studio 2017 SDK. To update one or more dependencies, use the following steps:
git clean -xfd
to make sure that there are only source files around.paket/paket.exe update
to update the dependenciesbuild
to build and test everythingRELEASE_NOTES.md
!publish
from command line to upload NuGet package (if you have the rights)set APIKEY=...
.nuget\NuGet.exe push bin\*.nupkg %APIKEY% -Source https://www.nuget.org