Merge pull request #48 from thorek1/higher_order__moments

Higher order moments

.github/workflows/ci.yml

@@ -14,25 +14,19 @@ jobs:
       fail-fast: false
       matrix:
         version:
-          # - '1.8'
           - '1.9' #^1-0 and ^1.9-0 are not recognised
-          # - 'nightly' # fails for zygote
         os:
           - ubuntu-latest
-          - macOS-latest #slow
-          - windows-latest #slow
+          - macOS-latest
+          - windows-latest
         arch:
           - x64
-          # - aarch64 #not implemented in Github CI
         include:
           - os: ubuntu-latest
             prefix: xvfb-run
           - version: '1.8'
             os: ubuntu-latest
             arch: x64
-          #- version: '1.8'
-          #  os: windows-latest
-          #  arch: x64
           - version: '^1.10.0-0'
             os: ubuntu-latest
             arch: x64
@@ -42,18 +36,8 @@ jobs:
             arch: x64
             allow_failure: true
     steps:
-      - uses: actions/checkout@v3
-      #- uses: uraimo/[email protected]
-      # - name: Set up QEMU
-      #   if: matrix.arch == 'aarch64'
-      #   uses: docker/setup-qemu-action@v2
-      # - name: Set up Docker Buildx
-      #   if: matrix.arch == 'aarch64'
-      #   uses: docker/setup-buildx-action@v2
-      # - name: Build and push
-      #   if: matrix.arch == 'aarch64'
-      #   uses: docker/build-push-action@v3
-      - uses: julia-actions/setup-julia@latest
+      - uses: actions/checkout@v4
+      - uses: julia-actions/setup-julia@v1
         with:
           version: ${{ matrix.version }}
           arch: ${{ matrix.arch }}
@@ -70,12 +54,11 @@ jobs:
             ${{ runner.os }}-test-${{ env.cache-name }}-
             ${{ runner.os }}-test-
             ${{ runner.os }}-
-      # - run: pip3 install sympy
-      - uses: julia-actions/julia-buildpkg@latest
-      - uses: julia-actions/julia-runtest@latest
+      - uses: julia-actions/julia-buildpkg@v1
+      - uses: julia-actions/julia-runtest@v1
         with:
           prefix: ${{ matrix.prefix }}  # for `xvfb-run`
-      - uses: julia-actions/julia-processcoverage@latest
+      - uses: julia-actions/julia-processcoverage@v1
       - uses: codecov/codecov-action@v3
         with:
           token: ${{ secrets.CODECOV_TOKEN }}

Project.toml

@@ -1,7 +1,7 @@
 name = "MacroModelling"
 uuid = "687ffad2-3618-405e-ac50-e0f7b9c75e44"
 authors = ["Thore Kockerols <[email protected]>"]
-version = "0.1.28"
+version = "0.1.29"
 
 [deps]
 AbstractDifferentiation = "c29ec348-61ec-40c8-8164-b8c60e9d9f3d"
@@ -31,6 +31,7 @@ SpeedMapping = "f1835b91-879b-4a3f-a438-e4baacf14412"
 Subscripts = "2b7f82d5-8785-4f63-971e-f18ddbeb808e"
 SymPyPythonCall = "bc8888f7-b21e-4b7c-a06a-5d9c9496438c"
 Symbolics = "0c5d862f-8b57-4792-8d23-62f2024744c7"
+ThreadedSparseArrays = "59d54670-b8ac-4d81-ab7a-bb56233e17ab"
 
 [weakdeps]
 StatsPlots = "f3b207a7-027a-5e70-b257-86293d7955fd"
@@ -61,6 +62,7 @@ SpeedMapping = "^0.3"
 Subscripts = "^0.1"
 SymPyPythonCall = "^0.1.1"
 Symbolics = "^5"
+ThreadedSparseArrays = "^0.2"
 julia = "1.8"
 
 [extras]

README.md

@@ -9,18 +9,18 @@
 
 **Author: Thore Kockerols (@thorek1)**
 
-`MacroModelling.jl` is a package for developing and solving dynamic stochastic general equilibrium (DSGE) models. The package provides functions for creating, calibrating, simulating and estimating discrete-time DSGE models. These kind of models are typically used to describe the behaviour of a macroeconomy and are particularly suited for counterfactual analysis (economic policy evaluation) and exploring/quantifying specific mechanisms (academic research). These models are difficult to work with because they consist of a nonlinear system of equations describing a stochastic control problem.
+`MacroModelling.jl` is a Julia package for developing and solving dynamic stochastic general equilibrium (DSGE) models. These kinds of models describe the behavior of a macroeconomy and are particularly suited for counterfactual analysis (economic policy evaluation) and exploring / quantifying specific mechanisms (academic research). Due to the complexity of these models, efficient numerical tools are required, as analytical solutions are often unavailable. `MacroModelling.jl` serves as a tool for handling the complexities involved, such as forward-looking expectations, nonlinearity, and high dimensionality.
 
-The goal of `MacroModelling.jl` is to reduce coding time and speed up model development.
+The goal of this package is to reduce coding time and speed up model development by providing functions for working with discrete-time DSGE models. The user-friendly syntax, automatic variable declaration, and effective steady state solver facilitates fast prototyping of models. The target audience for the package includes central bankers, regulators, graduate students, and others working in academia with an interest in DSGE modelling.
 
 As of now the package can:
 
 - parse a model written with user friendly syntax (variables are followed by time indices `...[2], [1], [0], [-1], [-2]...`, or `[x]` for shocks)
 - (tries to) solve the model only knowing the model equations and parameter values (no steady state file needed)
-- calculate first, second, and third order perturbation solutions using (forward or reverse-mode) automatic differentiation (AD)
+- calculate first, second, and third order (pruned) perturbation solutions using (forward or reverse-mode) automatic differentiation (AD)
 - calculate (generalised) impulse response functions, simulate the model, or do conditional forecasts
 - calibrate parameters using (non stochastic) steady state relationships
-- match model moments
+- match model moments (also for pruned higher order solutions)
 - estimate the model on data (Kalman filter using first order perturbation)
 - **differentiate** (forward AD) the model solution, Kalman filter loglikelihood (reverse-mode AD), model moments, steady state, **with respect to the parameters**
 
@@ -99,6 +99,7 @@ The package contains the following models in the `models` folder:
 **Automatic declaration of variables and parameters**|yes|||||||||||||
 **Derivatives (Automatic Differentiation) wrt parameters**|yes|||||||||yes - for all 1st, 2nd order perturbation solution related output *if user supplied steady state equations*|||
 **Perturbation solution order**|1, 2, 3|k|1 to 5|1|1|1|1|1, 2, 3|1, 2, 3|1, 2|1||1 to 5|
+**Pruning**|yes|yes|yes|||||||yes||||
 **Automatic derivation of first order conditions**|||||||||||yes||
 **Handles occasionally binding constraints**||yes|yes|||yes||yes|yes|||yes||
 **Global solution**||||||||yes|yes|||yes||
@@ -108,6 +109,7 @@ The package contains the following models in the `models` folder:
 **Timing convention**|end-of-period|end-of-period|end-of-period|end-of-period|end-of-period||end-of-period|start-of-period|end-of-period|start-of-period|end-of-period|start-of-period|start-of-period|
 
 ## Bibliography
+
 Andreasen, M. M., Fernández-Villaverde, J., and Rubio-Ramírez, J. F.  (2018), "The pruned state-space system for non-linear DSGE models: Theory and empirical applications.", The Review of Economic Studies, 85.1, p. 1-49.
 
 Durbin, J, and Koopman, S. J. (2012), "Time Series Analysis by State Space Methods, 2nd edn", Oxford University Press.

benchmark/SW07_estimation.jl

@@ -386,7 +386,7 @@ function calculate_kalman_filter_loglikelihoods(𝓂::ℳ, data::AbstractArray{F
 
     data_in_deviations = collect(data(observables)) .- SS_and_pars[obs_indices]
 
-	∇₁ = calculate_jacobian(parameters, SS_and_pars, 𝓂)
+	∇₁ = calculate_jacobian(parameters, SS_and_pars, 𝓂) |> Matrix
 
     sol = calculate_first_order_solution(∇₁; T = 𝓂.timings)
 
@@ -485,7 +485,7 @@ function calculate_kalman_filter_loglikelihoods(𝓂::ℳ, data::AbstractArray{F
 
     data_in_deviations = collect(data(observables)) .- SS_and_pars[obs_indices]
 
-	∇₁ = calculate_jacobian(parameters, SS_and_pars, 𝓂)
+	∇₁ = calculate_jacobian(parameters, SS_and_pars, 𝓂) |> Matrix
 
     sol = calculate_first_order_solution(∇₁; T = 𝓂.timings)
 

docs/make.jl

@@ -14,6 +14,7 @@ makedocs(
             "Installation" => "tutorials/install.md",
             "Write your first simple model - RBC" => "tutorials/rbc.md",
             "Work with a more complex model - Smets and Wouters (2003)" => "tutorials/sw03.md",
+            "Calibration / method of moments (for higher order perturbation solutions) - Gali (2015)" => "tutorials/calibration.md",
             "Estimate a model using gradient based samplers - Schorfheide (2000)" => "tutorials/estimation.md",
         ],
         "How-to guides" => [

docs/src/index.md

@@ -2,15 +2,15 @@
 
 **Author: Thore Kockerols (@thorek1)**
 
-`MacroModelling.jl` is a package for developing and solving dynamic stochastic general equilibrium (DSGE) models. The package provides functions for creating, calibrating, simulating and estimating discrete-time DSGE models. These kind of models are typically used to describe the behaviour of a macroeconomy and are particularly suited for counterfactual analysis (economic policy evaluation) and exploring/quantifying specific mechanisms (academic research). These models are difficult to work with because they consist of a nonlinear system of equations describing a stochastic control problem.
+`MacroModelling.jl` is a Julia package for developing and solving dynamic stochastic general equilibrium (DSGE) models. These kinds of models describe the behavior of a macroeconomy and are particularly suited for counterfactual analysis (economic policy evaluation) and exploring / quantifying specific mechanisms (academic research). Due to the complexity of these models, efficient numerical tools are required, as analytical solutions are often unavailable. `MacroModelling.jl` serves as a tool for handling the complexities involved, such as forward-looking expectations, nonlinearity, and high dimensionality.
 
-The goal of `MacroModelling.jl` is to reduce coding time and speed up model development.
+The goal of this package is to reduce coding time and speed up model development by providing functions for working with discrete-time DSGE models. The user-friendly syntax, automatic variable declaration, and effective steady state solver facilitates fast prototyping of models. The target audience for the package includes central bankers, regulators, graduate students, and others working in academia with an interest in DSGE modelling.
 
 As of now the package can:
 
 - parse a model written with user friendly syntax (variables are followed by time indices `...[2], [1], [0], [-1], [-2]...`, or `[x]` for shocks)
 - (tries to) solve the model only knowing the model equations and parameter values (no steady state file needed)
-- calculate first, second, and third order perturbation solutions using (forward or reverse-mode) automatic differentiation (AD)
+- calculate first, second, and third order (pruned) perturbation solutions using (forward or reverse-mode) automatic differentiation (AD)
 - calculate (generalised) impulse response functions, simulate the model, or do conditional forecasts
 - calibrate parameters using (non stochastic) steady state relationships
 - match model moments
@@ -54,6 +54,7 @@ The package contains the following models in the `models` folder:
 **Automatic declaration of variables and parameters**|yes|||||||||||||
 **Derivatives (Automatic Differentiation) wrt parameters**|yes|||||||||yes - for all 1st, 2nd order perturbation solution related output *if user supplied steady state equations*|||
 **Perturbation solution order**|1, 2, 3|k|1 to 5|1|1|1|1|1, 2, 3|1, 2, 3|1, 2|1||1 to 5|
+**Pruning**|yes|yes|yes|||||||yes||||
 **Automatic derivation of first order conditions**|||||||||||yes||
 **Handles occasionally binding constraints**||yes|yes|||yes||yes|yes|||yes||
 **Global solution**||||||||yes|yes|||yes||
@@ -63,6 +64,7 @@ The package contains the following models in the `models` folder:
 **Timing convention**|end-of-period|end-of-period|end-of-period|end-of-period|end-of-period||end-of-period|start-of-period|end-of-period|start-of-period|end-of-period|start-of-period|start-of-period|
 
 ## Bibliography
+
 Andreasen, M. M., Fernández-Villaverde, J., and Rubio-Ramírez, J. F.  (2018), "The pruned state-space system for non-linear DSGE models: Theory and empirical applications.", The Review of Economic Studies, 85.1, p. 1-49.
 
 Durbin, J, and Koopman, S. J. (2012), "Time Series Analysis by State Space Methods, 2nd edn", Oxford University Press.