Skip to content

Commit

Permalink
Fix docs and update cache action (#49)
Browse files Browse the repository at this point in the history
* Fix for Documenter pr2388

* Update action cache v4

* Specify using
  • Loading branch information
JeffFessler authored Jan 22, 2024
1 parent 4e7558c commit d6343fa
Show file tree
Hide file tree
Showing 7 changed files with 42 additions and 29 deletions.
2 changes: 1 addition & 1 deletion .github/workflows/Documentation.yml
Original file line number Diff line number Diff line change
Expand Up @@ -21,7 +21,7 @@ jobs:
with:
version: '1'
- name: CacheArtifacts
uses: actions/cache@v3
uses: actions/cache@v4
env:
cache-name: cache-artifacts
with:
Expand Down
9 changes: 5 additions & 4 deletions docs/lit/examples/1-overview.jl
Original file line number Diff line number Diff line change
Expand Up @@ -11,7 +11,8 @@ This page gives an overview of the Julia package

# Packages needed here.

using SPECTrecon
using SPECTrecon: plan_psf, psf_gauss, SPECTplan
using SPECTrecon: project, project!, backproject, backproject!
using MIRTjim: jim, prompt
using LinearAlgebra: mul!
using LinearMapsAA: LinearMapAA
Expand Down Expand Up @@ -218,10 +219,10 @@ mul!(tmp, A', views)
The pixel dimensions `deltas` can (and should!) be values with units.
Here is an example ... (todo)
=#
#using UnitfulRecipes
#using Unitful: mm
using UnitfulRecipes
using Unitful: mm
=#


# ## Projection view animation
Expand Down
2 changes: 1 addition & 1 deletion docs/lit/examples/2-rotate.jl
Original file line number Diff line number Diff line change
Expand Up @@ -11,7 +11,7 @@ This page explains the image rotation portion of the Julia package

# Packages needed here.

using SPECTrecon
using SPECTrecon: plan_rotate, imrotate!, imrotate_adj!
using MIRTjim: jim, prompt
using Plots: scatter, scatter!, plot!, default
default(markerstrokecolor=:auto, markersize=3)
Expand Down
2 changes: 1 addition & 1 deletion docs/lit/examples/3-psf.jl
Original file line number Diff line number Diff line change
Expand Up @@ -11,7 +11,7 @@ This page explains the PSF portion of the Julia package

# Packages needed here.

using SPECTrecon
using SPECTrecon: psf_gauss, plan_psf, fft_conv!, fft_conv_adj!
using MIRTjim: jim, prompt
using Plots: scatter, scatter!, plot!, default
default(markerstrokecolor=:auto, markersize=3)
Expand Down
2 changes: 1 addition & 1 deletion docs/lit/examples/4-mlem.jl
Original file line number Diff line number Diff line change
Expand Up @@ -11,7 +11,7 @@ This page illustrates ML-EM reconstruction with the Julia package

# Packages needed here.

using SPECTrecon
using SPECTrecon: SPECTplan, psf_gauss, project!, backproject!, mlem, mlem!
using MIRTjim: jim, prompt
using Plots: scatter, plot!, default; default(markerstrokecolor=:auto)

Expand Down
3 changes: 2 additions & 1 deletion docs/lit/examples/5-2d.jl
Original file line number Diff line number Diff line change
Expand Up @@ -12,7 +12,8 @@ using the Julia package

# Packages needed here.

using SPECTrecon
using SPECTrecon: SPECTplan, psf_gauss
using SPECTrecon: project, project!, backproject, backproject!
using MIRTjim: jim, prompt
using ImagePhantoms: shepp_logan, SheppLoganEmis
using LinearAlgebra: mul!
Expand Down
51 changes: 31 additions & 20 deletions docs/lit/examples/6-dl.jl
Original file line number Diff line number Diff line change
Expand Up @@ -219,26 +219,37 @@ end
# Initial loss
@show loss(xhat1, xtrue)

# ### Train the CNN
# Uncomment the following code to train!
## using Printf
## nepoch = 200
## for e = 1:nepoch
## @printf("epoch = %d, loss = %.2f\n", e, loss(xhat1, xtrue))
## ps = Flux.params(cnn)
## gs = gradient(ps) do
## loss(xhat1, xtrue) # we start with the 30 iteration EM reconstruction
## end
## opt = ADAMW(0.002)
## Flux.Optimise.update!(opt, ps, gs)
## end

# Uncomment to save your trained model.
## file = "../data/trained-cnn-example-6-dl.bson" # adjust path/name as needed
## @save file cnn

# Load the pre-trained model (uncomment if you save your own model).
## @load file cnn
#=
### Train the CNN
Uncomment the following code to train:
```
using Printf
nepoch = 200
for e in 1:nepoch
@printf("epoch = %d, loss = %.2f\n", e, loss(xhat1, xtrue))
ps = Flux.params(cnn)
gs = gradient(ps) do
loss(xhat1, xtrue) # we start with the 30 iteration EM reconstruction
end
opt = ADAMW(0.002)
Flux.Optimise.update!(opt, ps, gs)
end
```
=#

#=
Uncomment to save your trained model:
```
file = "../data/trained-cnn-example-6-dl.bson" # adjust path/name as needed
@save file cnn
```
Load the pre-trained model (uncomment if you save your own model):
```
@load file cnn
```
=#

#=
The code below here works fine when run via `include` from the REPL,
Expand Down

0 comments on commit d6343fa

Please sign in to comment.