Fixing the merge situation

.github/workflows/python-publish.yml

@@ -0,0 +1,31 @@
+# Sciope PYPI workflow
+# For more information see: https://help.github.com/en/actions/language-and-framework-guides/using-python-with-github-actions#publishing-to-package-registries
+
+name: Upload Python Package
+
+on:
+  release:
+    types: [created]
+
+jobs:
+  deploy:
+
+    runs-on: ubuntu-latest
+
+    steps:
+    - uses: actions/checkout@v2
+    - name: Set up Python
+      uses: actions/setup-python@v2
+      with:
+        python-version: '3.x'
+    - name: Install dependencies
+      run: |
+        python -m pip install --upgrade pip
+        pip install setuptools wheel twine
+    - name: Build and publish
+      env:
+        TWINE_USERNAME: __token__
+        TWINE_PASSWORD: ${{ secrets.PYPI }}
+      run: |
+        python setup.py sdist bdist_wheel
+        twine upload dist/*

.travis.yml

@@ -1,10 +1,8 @@
 language: python
 python:
-  - "3.5"
   - "3.6"
-# command to install dependencies
-before_install:
-  - sudo apt-get install -y libxml2 libxml2-dev
+  - "3.7"
+  - "3.8"
 install:
   - python3 -m pip install gillespy2
   - pip install pytest
@@ -15,15 +13,9 @@ install:
   - pip install .
 # tests should run from the tests directory
 before_script:
-  - wget https://github.com/StochSS/StochKit/archive/master.zip
-  - unzip master.zip
-  - cd StochKit-master
-  - ./install.sh
-  - export STOCHKIT_HOME=$PWD
-  - cd ..
   - cd sciope/tests
 # command to run tests
-script: travis_wait 30 py.test  --cov-report term --cov
+script: pytest  --cov-report term --cov
 after_success:
   - codecov
   - cd ../../docs/ && make html
@@ -33,7 +25,7 @@ deploy:
   skip-cleanup: true
   keep-history: true
   on:
-    branch: develop
+    branch: master
   github-token: $GITHUB_TOKEN
   local-dir: docs/build/html
   committer-from-gh: true

docs/source/index.rst

@@ -8,16 +8,16 @@
    :width: 100%
    :align: center
 
-Welcome to sciope's documentation!
+Welcome to Sciope's documentation!
 ===============================
 
 Scalable inference, optimization and parameter exploration (sciope)
 is a Python 3 package for performing machine learning-assisted likelihood-free inference and model
 exploration by large-scale parameter sweeps. It has been designed to simplify the data-driven workflows 
-so that users quickly can test and develop new machine learning-assisted approches to likelihood-free inference
+so that users quickly can test and develop new machine learning-assisted approaches to likelihood-free inference
 and model exploration.  
 
-Salient features and contributions of sciope include:
+Salient features and contributions of Sciope include:
 
 Systems:
 
@@ -38,7 +38,7 @@ Methodology:
 Stochastic Gene Regulatory Networks
 -----------------------------------
 Sciope has been designed for (but is not limited to) Stochastic Gene Regulatory Networks (GRN). 
-Sciope have built-in support and wrappers for `Gillespy2 <https://github.com/GillesPy2/GillesPy2>`_ 
+Sciope has built-in support and wrappers for `Gillespy2 <https://github.com/GillesPy2/GillesPy2>`_
 and is part of the development of next-generation `StochSS <https://stochss.org>`_.
 
 Likelihood-free inference
@@ -66,23 +66,23 @@ with a `Dask <https://dask.org>`_ backend to support massive parallelism on plat
 
 Installation
 ===============================
-You can install sciope with  ``pip``, or by installing from source.
+You can install Sciope with  ``pip``, or by installing from source.
 
 Pip
 ---
 
-This will install both sciope and other dependencies like NumPy, sklearn,
+This will install both Sciope and other dependencies like NumPy, sklearn,
 and so on that are necessary::
 
    pip install sciope
 
 Install from Source
 -------------------
 
-To install sciope from source, clone the repository from `github
-<https://github.com/sciope/sciope>`_::
+To install Sciope from source, clone the repository from `github
+<https://github.com/StochSS/sciope>`_::
 
-    git clone https://github.com/sciope/sciope.git
+    git clone https://github.com/StochSS/sciope.git
     cd sciope
     pip install .
 
@@ -162,7 +162,7 @@ Use Sciope's Gillespy2 wrapper to extract simulator and parameters
 
 Use Latin Hypercube design to generate points which will be sampled from during exploration, the points will
 be generated using distributed resources if we have a Dask client initialized (in this example just a local cluster).
-Generated points will be persited over the worker nodes (i.e no local memory would be used in case of a real cluster).
+Generated points will be persisted over the worker nodes (i.e no local memory would be used in case of a real cluster).
 Random points from the persisted collection can be gathered by calling :code:`lhc.draw(n_samples)`
 Here, we will also use TSFRESH minimal feature set as our summary statistics.
 
@@ -210,7 +210,7 @@ it is therefore required that you run in a jupyter notebook with an interactive
    :width: 100%
    :align: center
 
-Once at least a few points have been assigned a label, sciope has support for semi-supervised learning using label propagation where 
+Once at least a few points have been assigned a label, Sciope has support for semi-supervised learning using label propagation where
 we can infer the labels of unassigned points. This is a great way of filtering the vast amount of data according qualitative behaviour 
 and preferences.
 

examples/inference/MA2/ma2_inference.ipynb

@@ -46,6 +46,8 @@
    "outputs": [],
    "source": [
     "# Set up simulation\n",
+    "# We use 10,000 samples in the training dataset to be able to run the notebook quickly, which is comparitively few\n",
+    "# For accurate inference, the number should be 1 or 2 orders of magnitude more.\n",
     "sim = simulate\n",
     "true_param = [0.6, 0.2]  # true \\theta moving average2\n",
     "\n",
@@ -59,7 +61,7 @@
     "validation_ts = np.expand_dims(np.array([simulate(p, n=100) for p in validation_thetas]), 2)\n",
     "\n",
     "test_thetas = np.array(prior(n=10000))\n",
-    "test_ts = np.expand_dims(np.array([simulate(p, n=100) for p in validation_thetas]), 2)\n",
+    "test_ts = np.expand_dims(np.array([simulate(p, n=100) for p in test_thetas]), 2)\n",
     "\n",
     "abc_trial_thetas = np.array(prior(n=30000))\n",
     "abc_trial_ts = np.expand_dims(np.array([simulate(p, n=100) for p in abc_trial_thetas]), 2)"
@@ -233,9 +235,9 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "CNN test MAE = 0.7031869420803443, MAE at true point = 0.1392668604850769\n",
-      "DNN test MAE = 0.7157589187285183, MAE at true point = 0.2387733519077301\n",
-      "PEN test MAE = 0.7589048176817108, MAE at true point = 0.12148187458515167\n"
+      "CNN test MAE = 0.3804093829415728, MAE at true point = 0.571735179424286\n",
+      "DNN test MAE = 0.32691127034054135, MAE at true point = 0.2187902897596359\n",
+      "PEN test MAE = 0.19167063390147682, MAE at true point = 0.1489650845527649\n"
      ]
     }
    ],
@@ -323,8 +325,8 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "[[0.36439425 0.24292797]]\n",
-      "MAE upon comparison as a statistic = 0.1392668604850769\n"
+      "[[1.21066892 0.73280144]]\n",
+      "MAE upon comparison as a statistic = 0.571735179424286\n"
      ]
     }
    ],
@@ -420,7 +422,7 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "ABC mean inference error = 1.1916615337995151\n"
+      "ABC mean inference error = 1.172768172952754\n"
      ]
     }
    ],