Skip to content

Instantly share code, notes, and snippets.

Avatar
💭
Maintaining BigML bindings

mmerce

💭
Maintaining BigML bindings
View GitHub Profile
View model-evaluate.json
{
"name": "Model and Evaluate",
"description": "It creates an evaluation from an 80/20% split",
"inputs": [
{
"name": "dataset-id",
"type": "dataset-id",
"description": "Dataset to be modeled"
}
],
View model-evaluate.json
{
"name": "Model and Evaluate",
"description": "It creates an evaluation from an 80/20% split",
"inputs": [
{
"name": "dataset-id",
"type": "dataset-id",
"description": "Dataset to be modeled"
}
],
@mmerce
mmerce / 1-click-batchprediction.json
Last active Jul 23, 2018
1-click batch prediction with fusion
View 1-click-batchprediction.json
{
"name": "1-click batch prediction from fusion",
"description": "This script creates a batch prediction from a user-given source and fusion ID",
"inputs": [
{
"default": "",
"description": "Source for inputs",
"name": "test-source",
"type": "source-id"
},
@mmerce
mmerce / clean_anomaly.js
Created Jun 21, 2018
Creating a clean anomaly detector
View clean_anomaly.js
/*
* Creating a clean anomaly detector
*/
var bigml = require('bigml');
// Simplest anomaly detector
new bigml.Source().create('sales.csv', undefined, function(error, source) {
new bigml.Dataset().create(source, undefined, function(error, dataset) {
new bigml.Anomaly().create(dataset, undefined, function(error, anomaly) {
@mmerce
mmerce / move_to_project.json
Last active Apr 1, 2019
Moving from one project to another
View move_to_project.json
{"name": "Move from projects to project",
"description": "Moves the non-script resources from a list of existing projects to a single destination project",
"inputs": [
{
"default": "",
"description": "Destination project ID",
"name": "destination",
"type": "project-id"
},
{
@mmerce
mmerce / bigmler_model.sh
Last active Jan 16, 2018
Basic BigMLer model command
View bigmler_model.sh
bigmler --train data/diabetes_sample.csv \
--tag best_diabetes
--output-dir ./initial_model
@mmerce
mmerce / bigmler_retrain_id.sh
Created Jan 16, 2018
Post example: bigmler retrain by ID
View bigmler_retrain_id.sh
bigmler retrain --add diabetes_new_data.csv \
--id cluster/5a186f1d92527304c200077b \
--output-dir accumulative_cluster
View bigmler_retrain_window.sh
bigmler retrain --add diabetes_12.csv \
--window-size 3 \
--ensemble-tag best_diabetes \
--output-dir windowed_retrain
@mmerce
mmerce / retrain_add.sh
Created Jan 16, 2018
BigMLer retrain: add data
View retrain_add.sh
bigmler retrain --add diabetes_new_data.csv \
--model-tag best_diabetes \
--output-dir accumulative_retrain
@mmerce
mmerce / README.md
Last active Oct 6, 2017
Local batch centroid using Python bindings
View README.md

Python example: local batch centroid

Using the Python bindings and its local cluster object to find the centroids for a CSV file.

Usage:

cat input.csv | python local_batch_centroid.py > centroids.csv