noobvehicle.blogg.se

1. #CONDA INSTALL PACKAGE LOCAL HOW TO#
2. #CONDA INSTALL PACKAGE LOCAL INSTALL#
3. #CONDA INSTALL PACKAGE LOCAL UPGRADE#
4. #CONDA INSTALL PACKAGE LOCAL FULL#

The custom packages have a built-in test suite that can easily be run using pytest.For others who run into this situation, I found this to be the most straightforward solution: More info are available here Testing Suite The project implements three custom packages. Please look at the pre-requisites for more information about having a Google Cloud Storage Account. If the user wants to use Google Cloud Storage, will need to authenticate it like Google Earth Engine above. The notebooks, therefore, have been setup to accomodate either. The user may prefer to use Google Cloud Storage over Google Drive in some paet of the workflow (keep in mind tha the use of Google Drive is mandatory). Authenticating Google Cloud Storage (Optional) Please look at the pre-requisites above if you don't have a Google Earth Engine account yet. The process is identical to mounting a Google Drive above. When running the cells to import the required packages, the notebook will show a link and prompt the user to click on it (the link will be opened on a new page), login to the Gmail account (please look at the pre-requisites above if you don't have an account yet) and copy the one-time key that appears on the page into the empty box that appears within the Google Colab notebook as shown in the picture below.Įvery notebook requires that the user authenticates Google Earth Engine. The notebooks require the user to mount a target google drive for the workflow to function properly. If the project ever becomes public, the token will be removed.

#CONDA INSTALL PACKAGE LOCAL FULL#

The reason for providing one is because the current repo is private, and as such providing the Token to others (in this case the markers of the project) is the only way to get full access to the repo. This action is not generally advisable because it provides full access and to the repo and full ability to push changes to it to whomever possesses it.

#CONDA INSTALL PACKAGE LOCAL INSTALL#

This is necessary to access the requirements.txt file, that will be used to install all the necessary pakcages, inlcluding three custom packages (please see section Custom Packages below for details). Right at the start of each notebook, the user is prompted to clone the current github repository. Nonethelss, if the aim of the user is to take advantage of the full monitoring framework, then it is necessary to follow the numbering of the notebooks (1 to 3) as showed in the table above. The user may decide which notebook to use according to needs. The notebooks have been designed to work as standalones and each has a speficic task as described above. Import TFRecords and convert them into datasets ready to be fed into Neural Networks (NNs) use Keras API to train newly generated or pre-trained NNsĪcquisition of satellite imagery of a small target area using GEE export the image as TFRecord patches load NNs models to make predictions of the target image The monitoring framework has been split into three Google Colab notebooks a follow: NotebookĪcquisition of satellite imagery from the Google Servers using GEE classification of a target geographical area export of the classified image as TensorFlow (TF) records (TFRecords) patches of user-defined size (pixels)

#CONDA INSTALL PACKAGE LOCAL UPGRADE#

The first step is to upgrade the current version of pip by typing: USE THIS SECTION ONLY IF WANTING TO EXECUTE THE PROJECT ON A LOCAL MACHINE - NOTE THIS IS NOT ADVISED UNLESS POSSESSING A POWERFUL COMPUTER WITH LARGE GPU/ TPU MEMORY. Please see this guide for further details: Nevertheless, given that the aim of the project is to provide a fully-functional online platform, this was mainly developed for internet browser use only and the use on local machines is not adviced unless possessing a powerful computer with large memory GPU/ TPU Installation on local machine (Optional)

If the user wishes to do so, however, will need to enable Google Drive API access. Please Note: Google Collaboratory is built on Jupyter Notebooks and as such, the notebooks that it generates and that are saved on Google Drive may be downloaded and used on a local machine.

Nonetheless, the user may wish to use Google Cloud Storage instead by setting up a google storage bucket using the following link: The frameowrk is setup to rely on Google Drive as the main Cloud Storage.

#CONDA INSTALL PACKAGE LOCAL HOW TO#

Have a Google Earth Engine account -> please see instructions how to get one here:.Have a Gmail account -> please see instructions to create one here:.The project is completely cloud-based and it uses Google Collaboratoryas the core platform. Built Withįollow the instructions below to correctly run the monitoring framework. The aim of the project is to build a cloud-based mangroves monitoring framework using Google Collaboratory, Google Earth Engine and Google Storage and using Convolutional Neural Networks for the classification of landcover types. Authenticating Google Cloud Storage (Optional).