{"id":14977841,"url":"https://github.com/andresberejnoi/netbuilder","last_synced_at":"2025-05-06T17:34:37.636Z","repository":{"id":57445615,"uuid":"52160235","full_name":"andresberejnoi/NetBuilder","owner":"andresberejnoi","description":"Create fully connected neural networks that can be trained in just a few lines of code","archived":false,"fork":false,"pushed_at":"2022-05-05T00:02:59.000Z","size":414,"stargazers_count":2,"open_issues_count":0,"forks_count":3,"subscribers_count":2,"default_branch":"master","last_synced_at":"2024-11-12T22:09:44.719Z","etag":null,"topics":["educational-project","feedforward-neural-network","machine-learning","netbuilder","neural-network","neuralnetwork","numpy","python","python3"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/andresberejnoi.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null}},"created_at":"2016-02-20T15:33:13.000Z","updated_at":"2022-05-05T00:03:02.000Z","dependencies_parsed_at":"2022-09-26T16:30:59.994Z","dependency_job_id":null,"html_url":"https://github.com/andresberejnoi/NetBuilder","commit_stats":null,"previous_names":[],"tags_count":7,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/andresberejnoi%2FNetBuilder","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/andresberejnoi%2FNetBuilder/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/andresberejnoi%2FNetBuilder/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/andresberejnoi%2FNetBuilder/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/andresberejnoi","download_url":"https://codeload.github.com/andresberejnoi/NetBuilder/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":224517393,"owners_count":17324407,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["educational-project","feedforward-neural-network","machine-learning","netbuilder","neural-network","neuralnetwork","numpy","python","python3"],"created_at":"2024-09-24T13:56:25.296Z","updated_at":"2024-11-13T20:08:44.029Z","avatar_url":"https://github.com/andresberejnoi.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"#          NetBuilder\nThe neural network class is in NeuralNet.py.\nIt allows to easily create fully connected feedforward networks of\nany size allowed by available memory. It uses numpy arrays as the primary\ndata structure for the weight matrices.\nWith this package you can create deep neural networks very quickly.\n\n## Installation\nPackage can be installed with pip:\n\n```sh\npip install netbuilder\n```\n## How to Use\nThe project's documentation was built using Sphinx and stored at readthedocs.org but for some reason it stopped working there. I need to rebuild it and that will happen when I find the time. In the meantime, the same documentation is spread throughout the source code as docstrings. I will include a small piece below:\n\nTo use the package, it has to be imported first like:\n```\n  \u003e\u003e\u003e import netbuilder\n  or\n  \u003e\u003e\u003e import netbuilder as nb\n\n  With the following lines, you can create a neural network for\n  a binary gate:\n\n  \u003e\u003e\u003e net = nb.Network()\n  \u003e\u003e\u003e net.init(topology=[2,1])\n\n  The  first line above will create a `Network` object. The parameters\n  of the network are not defined yet. The second line tells it to initialize\n  weights for a shape of two input nodes for the first layer and one output\n  node at the final layer.\n  To create hidden layers, just add them to the topology parameter when\n  initializing the network:\n\n  \u003e\u003e\u003e net = nb.Network()\n  \u003e\u003e\u003e net.init(topology=[2,5,5,1])\n\n  The above lines will create a `Network` object with 4 layers: one input layer\n  with 2 nodes, two hidden layers with 5 nodes each, and an output layer with\n  one node.\n  To perform a feedforward propagation an input array is needed. If the array\n  is a numpy array with shape [number of samples x number of features], then\n  the `feedfoward` method can be used:\n\n  \u003e\u003e\u003e x = numpy.array([[0,1]])\n  \u003e\u003e\u003e net.feedforward(x)\n  array([[ 0.82683518]])\n\n  Note above that the following format for x will cause an error because the\n  shape is (,2) when it should be (1,2):\n\n  \u003e\u003e\u003e x = numpy.array([0,1])\n\n  The method `predict` is available for quick testing without worry about the\n  format of the input array:\n\n  \u003e\u003e\u003e x = [0,1]\n  \u003e\u003e\u003e net.predict(x)\n  array([[ 0.82683518]])\n```\n\nAnd documentation built with Sphinx can be found at (not working for now):\nhttp://netbuilder.readthedocs.io/en/latest/\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fandresberejnoi%2Fnetbuilder","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fandresberejnoi%2Fnetbuilder","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fandresberejnoi%2Fnetbuilder/lists"}