Dropbox Paper

Building a Fully Functional Javascript NeuroApp

Now that we've got the basics out of the way, we can start to build something that actually uses the FFBO system. We will create a pure JavaScript and HTML web page that is aware of the FFBO resources and interacts directly with the NeuroNLP and NeuroArch components.

This is a Work In Progress, see the Simple branch of the Skeleton_Interface.

Prerequisites

Processor Component Installed and Running

NeuroNLP Component Installed and Running

Neuroarch Component Installed and Running

What are we going to build?

A simple neuron viewer from start to finish

Project Direction

In order to complete this project, we will:

Connect to the FFBO project

Convert some Natural Language to the Neuroarch API

Query NeuroArch

Display the result TODO

Getting Started

We will build on the skeleton program, and first contact the NeuroNLP server

First, we need to make sure the FFBO servers are up and running.

We will then build on the skeleton app, which you can get with git with #todo

git clone https://github.com/fruitflybrain/ffbo.skeleton_interface

Connecting with NeuroNLP

In this example, we will automatically start the query processing when we receive the server list.

First, we need to understand how to directly address the NLP server. In the server list, we have the keys of the Types of server, (ie nlp,na,nk...). Inside each object, we have a dictionary of servers of that type. Likely we will only have the one, but the FFBO supports multiple servers running for both improved functionality and redundancy.

So in this case, we will first get the first NLP server id, like so:

var server_id = Object.keys(server_list['nlp'])[0]

Then, we will have to write a function to neatly do the conversion for us. Before we do that, we will write the code to call this function, which we will call nlp_query:

var query_result = nlp_query(session,server_id);

We will pass in the server ID that we have found, and also pass along the session object, to enable the external function to access the current FFBO session.

Now, we can start to write the function, outside of the on_connect class:

// define a function to retrieve the neuron

function nlp_query(session,server_id){

var query = "Show L1 neurons in the Lamina"