when you run an analysis in edgeful AI, the result comes back as a structured data table. each row represents a time period — a specific date or day — and each column represents a metric from the report(s) you selected.

the table is your primary output. everything you'll want to explore, question, or act on starts here.

every data table — regardless of which report you run — includes 2 columns that are always present:

these are your anchors. they let you spot patterns by weekday, identify specific dates you want to ask about, and provide the time context for every other metric in the row.

beyond Date and Day, every other column in the table is specific to the report you selected. the columns vary by report — an IB analysis will produce different metrics than an Opening Candle Continuation analysis or a gap analysis.

these columns contain the actual data for each report: the numbers, rates, and readings that the analysis is built on. what they mean depends on which report is in the table.

if you're not sure what a specific column represents, you can ask the AI directly — just reference the column name and ask it to explain. that's covered below.

if you run a multi-report analysis, the data table uses colour coding to separate the different reports from each other — not to indicate direction or performance.

for example, if you run IB Standard and Opening Candle Continuation together, the IB columns will appear in one colour and the OCC columns in a different colour. this makes it easy to see at a glance which metrics belong to which report when everything is side by side in one table.

the colours are purely organisational. a green-coded column doesn't mean bullish — it just means "these columns belong to report X." don't read directional meaning into the colours in the data table.

if a cell in the data table shows a dash (—), it means there's no data available for that particular date and metric combination.

this can happen for a few reasons — the instrument wasn't trading on that date, there's not enough historical data to produce a result for that specific condition, or the report metric doesn't have a value for that row.

a dash isn't an error. it just means the data doesn't exist for that cell. rows with many dashes are generally less useful for pattern recognition — you can ask the AI to filter those out or focus only on dates with complete data.

this is one of the most useful things to know about working with the data table: you can reference column labels directly in your follow-up questions, and the AI will know exactly which metric you're asking about.

column labels work for 2 things:

understanding a metric
if a column name isn't immediately clear, ask about it by name. the AI will explain what that metric is measuring, how it's calculated, and what to look for in the values.

analysing specific data
you can ask the AI to filter, sort, or draw conclusions from a specific column. for example, asking the AI to identify which dates had the highest reading in a specific column, or to tell you how a particular metric has trended over the date range, gives you a targeted analysis of exactly the data point you care about.

the more specific your follow-up question, the better the response. using the exact column label from the table removes any ambiguity — the AI knows precisely what you're asking about and can work with the data directly.