{ "info": { "author": "Louis Moresi", "author_email": "louis.moresi@unimelb.edu.au", "bugtrack_url": null, "classifiers": [], "description": "Miller & Moresi - Alaska Moho Model with Notebooks\n====================================\n\n---\n\n![Image showing map of results](https://www.dropbox.com/s/5s6uk6m3dlg5ysq/MohoSurfaceGradient-ClusteredGrids.png?raw=1)\n\n_Preferred model, gradient and a comparison to heat flow data. The maps were\nproduced using the `cartopy` package. Instructions for reproducing these\nmaps are in the notebooks supplied with this package_\n\n---\n\n\nThis package is a self-consistent packaging of the Miller & Moresi Alaska Moho model.\nThere are many ways to access this package.\n\nIt contains\n\n - This information\n - Scripts to install documentation and examples\n - Jupyter notebooks for manipulating the data\n - Jupyter notebooks for recreating our work\n\nA demo is available on [mybinder.org](http://mybinder.org) - open the notebook `A0-Index.ipynb`\nto begin. The 'current' version with any updates / errata is here:\n\n[![Binder](https://mybinder.org/badge.svg)](https://mybinder.org/v2/gh/lmoresi/miller-moho-binder/master)\n\nThis is exactly the published version (1.0), warts and all.\n\n[![Binder](https://mybinder.org/badge.svg)](https://mybinder.org/v2/gh/lmoresi/miller-moho-binder/publication)\n\n\nSite Map\n--------\n\nThe jupyter notebooks provided as examples are given below\n\n - A1: Raw data, convert and save\n - A2: Raw data - plot quality information\n - A3: Triangulating and interpolating raw data\n - A4: Plotting moho and moho slope\n - A5: Interactive 3D plot\n - A6: Convert Models to Regular XYZ grid\n\nThere are further notebooks that allow you to reproduce our work in creating the\nprefered model and these can be browsed in the `ModelConstruction` directory\n\n\n---\n\n![Image showing quality scores on a map](https://www.dropbox.com/s/n6y4c8h0eauvuhv/ErrorsAndScores.png?raw=1)\n\n_We constructed our model by fitting surfaces to the observed points and recording the capacity\nof the model to predict information at points not included in the model construction. The\nquality of the model at each point was given an integer score and this was used to weight information\nin the final fitting process_\n\n\n\nInstallation\n-----------\n\nThe package can be installed standalone using `pip`\nor it can be run through docker without needing specific installation.\n\nThese notebooks can be viewed with jupyter but will only run if all the software dependencies have\nbeen installed (see [Installation through `pip`](#Installation-through-pip) below.\n\nIn nearly every instance we recommend using the (self-contained) docker version.\nWe have provided some useful `bash` shortcuts that make the docker commands\neasier to remember (see [Installation through docker](#Installation-through-docker) below).\n\n\n\n### Installation through docker\n\n\nFirst it is necessary to install the free __docker, community edition__ from the [docker store](https://store.docker.com/search?offering=community&type=edition) for your platform.\n\n```bash\n# Download the image with the scripts and data\ndocker pull lmoresi/docker-miller-moho:latest\n```\n\nThat's it !\n\nTo test the installation, try the following\n\n#### Command line docker examples\n\nThis help message\n\n```bash\n# print help message (i.e. usage)\ndocker run --rm lmoresi/docker-miller-moho:latest help\n```\n\nInstall the bash helpers in the current directory\n```bash\n# print help message (i.e. usage)\ndocker run --rm lmoresi/docker-miller-moho:latest bash_utils > msmoho_bash_utils.sh\nsource msmoho_bash_utils.sh\n```\n\nInstall the documentation / scripts and notebooks in the current directory\n```bash\n# print help message (i.e. usage)\nsource msmoho_bash_utils.sh\nmsmoho-docker-sh install_examples\n```\n\nRun a local python script \n\n```bash\n# run my_script.py with python in the docker container\nsource msmoho_bash_utils.sh\nmsmoho-docker-sh my_script.py\n\n```\n\n### Installation through pip (or conda)\n\nThe installation of the python package is straightforward but\nthere is a dependency on numpy and on stripy which may require\nan installed fortran compiler. If this proves problematic, the\ndocker version may simply be the best choice.\n\n```bash\n #! /bin/env bash\n # install the main package\n # stripy and numpy are installed as dependencies by pip\n # but we can do this explicitly to manage versions and\n # check for errors\n\n pip install numpy\n pip install stripy\n\n # stripy and numpy\n # as they embed fortran and C packages\n # miller_alaskamoho_srl2018 itself is pure python\n\n pip install miller_alaskamoho_srl2018\n```\n\nThe following script tests the installation:\n\n```python\n #! /bin/env python\n import numpy as np\n try:\n import miller_alaskamoho_srl2018 as alaskamoho\n except ImportError:\n print (\"Problem importing the alaska moho package\")\n\n # Check the data files exist / can be read\n # [('lon', '