Rerun tutorials

markdown/00-introduction/00_introduction.md

@@ -62,7 +62,7 @@ data[1:5]
 ```
 
 ```
-5-element Vector{Bool}:
+5-element Array{Bool,1}:
  1
  0
  1
@@ -245,14 +245,14 @@ TuringTutorials.weave_file("00-introduction", "00_introduction.jmd")
 
 Computer Information:
 ```
-Julia Version 1.6.0
-Commit f9720dc2eb (2021-03-24 12:55 UTC)
+Julia Version 1.5.3
+Commit 788b2c77c1 (2020-11-09 13:37 UTC)
 Platform Info:
   OS: Linux (x86_64-pc-linux-gnu)
   CPU: Intel(R) Core(TM) i9-9900K CPU @ 3.60GHz
   WORD_SIZE: 64
   LIBM: libopenlibm
-  LLVM: libLLVM-11.0.1 (ORCJIT, skylake)
+  LLVM: libLLVM-9.0.1 (ORCJIT, skylake)
 Environment:
   JULIA_CMDSTAN_HOME = /home/cameron/stan/
   JULIA_NUM_THREADS = 16
@@ -262,12 +262,12 @@ Environment:
 Package Information:
 
 ```
-      Status `~/.julia/dev/TuringTutorials/tutorials/00-introduction/Project.toml`
-  [31c24e10] Distributions v0.24.15
-  [c7f686f2] MCMCChains v4.7.3
-  [91a5bcdd] Plots v1.11.1
+Status `~/.julia/dev/TuringTutorials/tutorials/00-introduction/Project.toml`
+  [31c24e10] Distributions v0.24.18
+  [c7f686f2] MCMCChains v4.9.0
+  [91a5bcdd] Plots v1.12.0
   [f3b207a7] StatsPlots v0.14.19
-  [fce5fe82] Turing v0.15.1
+  [fce5fe82] Turing v0.15.18
   [9a3f8284] Random
 
 ```

markdown/01-gaussian-mixture-model/01_gaussian-mixture-model.md

@@ -231,11 +231,11 @@ Package Information:
 
 ```
 Status `~/.julia/dev/TuringTutorials/tutorials/01-gaussian-mixture-model/Project.toml`
-  [31c24e10] Distributions v0.24.15
-  [c7f686f2] MCMCChains v4.7.3
-  [91a5bcdd] Plots v1.11.1
+  [31c24e10] Distributions v0.24.18
+  [c7f686f2] MCMCChains v4.9.0
+  [91a5bcdd] Plots v1.12.0
   [f3b207a7] StatsPlots v0.14.19
-  [fce5fe82] Turing v0.15.12
+  [fce5fe82] Turing v0.15.18
   [9a3f8284] Random
 
 ```

markdown/02-logistic-regression/02_logistic-regression.md

@@ -189,7 +189,7 @@ describe(chain)
 ```
 
 ```
-2-element Vector{MCMCChains.ChainDataFrame}:
+2-element Array{MCMCChains.ChainDataFrame,1}:
  Summary Statistics (4 x 7)
  Quantiles (4 x 6)
 ```
@@ -334,14 +334,14 @@ TuringTutorials.weave_file("02-logistic-regression", "02_logistic-regression.jmd
 
 Computer Information:
 ```
-Julia Version 1.6.0
-Commit f9720dc2eb (2021-03-24 12:55 UTC)
+Julia Version 1.5.3
+Commit 788b2c77c1 (2020-11-09 13:37 UTC)
 Platform Info:
   OS: Linux (x86_64-pc-linux-gnu)
   CPU: Intel(R) Core(TM) i9-9900K CPU @ 3.60GHz
   WORD_SIZE: 64
   LIBM: libopenlibm
-  LLVM: libLLVM-11.0.1 (ORCJIT, skylake)
+  LLVM: libLLVM-9.0.1 (ORCJIT, skylake)
 Environment:
   JULIA_CMDSTAN_HOME = /home/cameron/stan/
   JULIA_NUM_THREADS = 16
@@ -351,20 +351,20 @@ Environment:
 Package Information:
 
 ```
-      Status `~/.julia/dev/TuringTutorials/tutorials/02-logistic-regression/Project.toml`
-  [a93c6f00] DataFrames v0.22.6
+Status `~/.julia/dev/TuringTutorials/tutorials/02-logistic-regression/Project.toml`
+  [a93c6f00] DataFrames v1.0.1
   [b4f34e82] Distances v0.10.2
-  [31c24e10] Distributions v0.24.15
-  [5789e2e9] FileIO v1.6.5
+  [31c24e10] Distributions v0.24.18
+  [5789e2e9] FileIO v1.8.0
   [38e38edf] GLM v1.4.1
-  [c7f686f2] MCMCChains v4.7.3
-  [cc2ba9b6] MLDataUtils v0.5.3
-  [872c559c] NNlib v0.7.17
-  [91a5bcdd] Plots v1.11.1
-  [ce6b1742] RDatasets v0.7.4
-  [4c63d2b9] StatsFuns v0.9.7
+  [c7f686f2] MCMCChains v4.9.0
+  [cc2ba9b6] MLDataUtils v0.5.4
+  [872c559c] NNlib v0.7.19
+  [91a5bcdd] Plots v1.12.0
+  [ce6b1742] RDatasets v0.7.5
+  [4c63d2b9] StatsFuns v0.9.8
   [f3b207a7] StatsPlots v0.14.19
-  [fce5fe82] Turing v0.15.1
+  [fce5fe82] Turing v0.15.18
   [9a3f8284] Random
 
 ```

markdown/03-bayesian-neural-network/03_bayesian-neural-network.md

@@ -88,13 +88,13 @@ The end of this tutorial provides some code that can be used to generate more ge
 # Turn a vector into a set of weights and biases.
 function unpack(nn_params::AbstractVector)
     W₁ = reshape(nn_params[1:6], 3, 2);   
-    b₁ = reshape(nn_params[7:9], 3)
+    b₁ = nn_params[7:9]
 
     W₂ = reshape(nn_params[10:15], 2, 3); 
-    b₂ = reshape(nn_params[16:17], 2)
+    b₂ = nn_params[16:17]
 
     Wₒ = reshape(nn_params[18:19], 1, 2); 
-    bₒ = reshape(nn_params[20:20], 1)   
+    bₒ = nn_params[20:20]
     return W₁, b₁, W₂, b₂, Wₒ, bₒ
 end
 
@@ -146,20 +146,6 @@ N = 5000
 ch = sample(bayes_nn(hcat(xs...), ts), HMC(0.05, 4), N);
 ```
 
-```
-Error: MethodError: ReverseDiff.TrackedReal{Float64, Float64, ReverseDiff.T
-rackedArray{Float64, Float64, 2, Matrix{Float64}, Matrix{Float64}}}(::Forwa
-rdDiff.Dual{ForwardDiff.Tag{ReverseDiff.var"#105#107"{DataType, Tuple{}, Va
-l{(1,)}}, Float64}, Float64, 1}) is ambiguous. Candidates:
-  (T::Type{var"#s31"} where var"#s31"<:Real)(x::ForwardDiff.Dual) in Tracke
-r at /home/cameron/.julia/packages/Tracker/YNNTM/src/lib/real.jl:111
-  ReverseDiff.TrackedReal{V, D, O}(value) where {V, D, O} in ReverseDiff at
- /home/cameron/.julia/packages/ReverseDiff/iHmB4/src/tracked.jl:56
-Possible fix, define
-  ReverseDiff.TrackedReal{V, D, O}(::ForwardDiff.Dual) where {V, D, O}
-```
-
-
 
 
 
@@ -170,11 +156,6 @@ Now we extract the weights and biases from the sampled chain. We'll use these pr
 theta = MCMCChains.group(ch, :nn_params).value;
 ```
 
-```
-Error: UndefVarError: ch not defined
-```
-
-
 
 
 
@@ -200,11 +181,7 @@ Z = [nn_forward([x, y], theta[i, :])[1] for x=x_range, y=y_range]
 contour!(x_range, y_range, Z)
 ```
 
-```
-Error: UndefVarError: ch not defined
-```
-
-
+![](figures/03_bayesian-neural-network_7_1.png)
 
 
 
@@ -242,11 +219,7 @@ Z = [nn_predict([x, y], theta, n_end)[1] for x=x_range, y=y_range]
 contour!(x_range, y_range, Z)
 ```
 
-```
-Error: UndefVarError: theta not defined
-```
-
-
+![](figures/03_bayesian-neural-network_9_1.png)
 
 
 
@@ -263,11 +236,7 @@ anim = @gif for i=1:n_end
 end every 5
 ```
 
-```
-Error: UndefVarError: theta not defined
-```
-
-
+![](figures/03_bayesian-neural-network_10_1.gif)
 
 
 
@@ -294,20 +263,6 @@ advi = ADVI(10, 5_000)
 q_hat = vi(m, advi, q);
 ```
 
-```
-Error: MethodError: ReverseDiff.TrackedReal{Float64, Float64, ReverseDiff.T
-rackedArray{Float64, Float64, 2, Matrix{Float64}, Matrix{Float64}}}(::Forwa
-rdDiff.Dual{ForwardDiff.Tag{ReverseDiff.var"#105#107"{DataType, Tuple{}, Va
-l{(1,)}}, Float64}, Float64, 1}) is ambiguous. Candidates:
-  (T::Type{var"#s31"} where var"#s31"<:Real)(x::ForwardDiff.Dual) in Tracke
-r at /home/cameron/.julia/packages/Tracker/YNNTM/src/lib/real.jl:111
-  ReverseDiff.TrackedReal{V, D, O}(value) where {V, D, O} in ReverseDiff at
- /home/cameron/.julia/packages/ReverseDiff/iHmB4/src/tracked.jl:56
-Possible fix, define
-  ReverseDiff.TrackedReal{V, D, O}(::ForwardDiff.Dual) where {V, D, O}
-```
-
-
 
 ```julia
 samples = transpose(rand(q_hat, 5000))
@@ -317,11 +272,6 @@ ch_vi = Chains(reshape(samples, size(samples)..., 1), string.(MCMCChains.namesin
 theta = MCMCChains.group(ch_vi, :nn_params).value;
 ```
 
-```
-Error: UndefVarError: q_hat not defined
-```
-
-
 
 ```julia
 # Plot the average prediction.
@@ -334,11 +284,7 @@ Z = [nn_predict([x, y], theta, n_end)[1] for x=x_range, y=y_range]
 contour!(x_range, y_range, Z)
 ```
 
-```
-Error: UndefVarError: theta not defined
-```
-
-
+![](figures/03_bayesian-neural-network_13_1.png)
 
 
 
@@ -408,20 +354,6 @@ num_samples = 500
 ch2 = sample(bayes_nn_general(hcat(xs...), ts, network_shape, num_params), NUTS(0.65), num_samples);
 ```
 
-```
-Error: MethodError: ReverseDiff.TrackedReal{Float64, Float64, ReverseDiff.T
-rackedArray{Float64, Float64, 2, Matrix{Float64}, Matrix{Float64}}}(::Forwa
-rdDiff.Dual{ForwardDiff.Tag{ReverseDiff.var"#105#107"{DataType, Tuple{}, Va
-l{(1,)}}, Float64}, Float64, 1}) is ambiguous. Candidates:
-  (T::Type{var"#s31"} where var"#s31"<:Real)(x::ForwardDiff.Dual) in Tracke
-r at /home/cameron/.julia/packages/Tracker/YNNTM/src/lib/real.jl:111
-  ReverseDiff.TrackedReal{V, D, O}(value) where {V, D, O} in ReverseDiff at
- /home/cameron/.julia/packages/ReverseDiff/iHmB4/src/tracked.jl:56
-Possible fix, define
-  ReverseDiff.TrackedReal{V, D, O}(::ForwardDiff.Dual) where {V, D, O}
-```
-
-
 
 ```julia
 # This function makes predictions based on network shape.
@@ -440,11 +372,7 @@ Z = [nn_predict([x, y], params2, length(ch2), network_shape)[1] for x=x_range, y
 contour!(x_range, y_range, Z)
 ```
 
-```
-Error: UndefVarError: ch2 not defined
-```
-
-
+![](figures/03_bayesian-neural-network_15_1.png)
 
 
 
@@ -457,19 +385,19 @@ This has been an introduction to the applications of Turing and Flux in defining
 To locally run this tutorial, do the following commands:
 ```julia, eval = false
 using TuringTutorials
-TuringTutorials.weave_file("03-bayesian-neural-network", "03_bayesian-neural-network.jmd")
+TuringTutorials.weave_file("03-bayesian-neural-network/", "03_bayesian-neural-network.jmd")
 ```
 
 Computer Information:
 ```
-Julia Version 1.6.0
-Commit f9720dc2eb (2021-03-24 12:55 UTC)
+Julia Version 1.5.3
+Commit 788b2c77c1 (2020-11-09 13:37 UTC)
 Platform Info:
   OS: Linux (x86_64-pc-linux-gnu)
   CPU: Intel(R) Core(TM) i9-9900K CPU @ 3.60GHz
   WORD_SIZE: 64
   LIBM: libopenlibm
-  LLVM: libLLVM-11.0.1 (ORCJIT, skylake)
+  LLVM: libLLVM-9.0.1 (ORCJIT, skylake)
 Environment:
   JULIA_CMDSTAN_HOME = /home/cameron/stan/
   JULIA_NUM_THREADS = 16
@@ -479,13 +407,13 @@ Environment:
 Package Information:
 
 ```
-      Status `~/.julia/dev/TuringTutorials/tutorials/03-bayesian-neural-network/Project.toml`
-  [b5ca4192] AdvancedVI v0.1.1
-  [76274a88] Bijectors v0.8.13
-  [587475ba] Flux v0.12.0
-  [91a5bcdd] Plots v1.11.1
-  [37e2e3b7] ReverseDiff v1.7.0
-  [fce5fe82] Turing v0.15.1
+Status `~/.julia/dev/TuringTutorials/tutorials/03-bayesian-neural-network/Project.toml`
+  [b5ca4192] AdvancedVI v0.1.2
+  [76274a88] Bijectors v0.9.1
+  [587475ba] Flux v0.12.1
+  [91a5bcdd] Plots v1.12.0
+  [37e2e3b7] ReverseDiff v1.8.0
+  [fce5fe82] Turing v0.15.18
   [9a3f8284] Random
 
 ```