Nominal values are recorded in current money units (e.g. current US dollars). Real values adjust for inflation so you can compare purchasing power across years. SignalMap uses both: nominal for market prices, real or indexed series where studies need long-run comparability.

GDP is the total value of finished goods and services produced in an economy over a period. Studies use it as context for size (levels) and for composition (consumption, investment, oil rents) as shares of GDP.

Total GDP in dollars mixes real output, prices, and exchange rates, and larger populations mechanically raise many totals. Indexed charts (base year = 100) make relative growth easier to read on one axis; they do not replace per-capita or welfare measures when you care about living standards.

Final consumption is spending by households and government on goods and services. Gross capital formation is investment in capacity (equipment, construction, inventories). Charts that show each as a percent of GDP describe how the economy is split—not dollar size on the same scale as GDP in levels view.

Level charts can show current US dollars, constant-price (real) dollars, or illustrative conversions (e.g. toman using open-market FX). Indexed charts rescale each series to 100 in a base year so different units can be compared for relative change—not for reading absolute dollars on one axis.

A structural break is a time when a series visibly changes level or trend. On SignalMap, breaks are discussed descriptively alongside events—they are not automatic statistical tests and do not prove that an event caused the change.

The Gini summarizes how unequal a distribution is (often income), on a 0–100 scale. Higher means more concentration; lower means more equality. Cross-country charts use the same statistical definition so lines are comparable, but definitions and surveys still differ from lived experience.

Consumer price index (CPI) inflation as “percent change from a year ago” compares this month or year to the same period last year. It is a standard way to read how fast average consumer prices are rising.

The poverty headcount is the share of the population below an international poverty line defined by the World Bank. Different indicators use different thresholds (e.g. lower-middle-income line vs. a fixed line); they are not additive and are not the same as national poverty statistics.

Nominal wages are the statutory or reported pay in current currency. Real wages adjust for inflation (e.g. with CPI) so you can see whether pay buys more or less over time. That is the same “purchasing power” idea as real prices, applied to labor income.

Oil prices are benchmarks for crude sold on world markets. They signal energy costs and broader stress; SignalMap uses them as context series, not as proof of any single cause.

Brent and West Texas Intermediate (WTI) are two benchmark crudes. Brent is common for international pricing; WTI is a US benchmark. SignalMap’s oil studies use Brent unless noted otherwise.

USD/bbl is the price of one barrel of oil in US dollars. A barrel is about 159 litres—the standard unit for crude quotes.

Real oil adjusts nominal crude prices for inflation (e.g. with CPI) so long-run burden can be compared across decades. Nominal prices from the past look “cheap” in dollars; real prices show whether oil was relatively expensive in purchasing-power terms.

Gold (USD/oz) appears with oil in century-scale charts as a second stress indicator. Axes and frequencies differ from oil; use the pair for broad historical context, not as a single combined index.

Some charts flag days when the daily oil return is large compared to recent volatility—a simple “unusual move” filter for reading the series, not a full list of causes.

PPP converts currencies using what money buys domestically, not only market exchange rates. PPP-based “oil burden” studies multiply a world oil price by a PPP factor so the cost of a barrel is expressed in local purchasing power—useful for comparing Iran and Turkey with the same construction.

Oil rents are resource income measured as a share of GDP in national accounts. Production is how much crude is extracted (often in barrels per day). Exports are what is sold abroad—they can be far below production when domestic use or constraints bind. Studies separate these ideas on purpose.

A high world price does not mean a country can sell more crude. Sanctions, logistics, or quotas hit volumes. SignalMap’s export-capacity view combines price and estimated volume as a descriptive proxy—not realized government revenue.

Heavy reliance on one commodity for exports or budget revenue makes growth and public finance sensitive to price swings. That volatility is a reason charts combine prices with volumes or wider macro panels—not a verdict on policy by itself.

Network charts show who trades crude with whom: nodes are countries, links are flows. They help visualize concentration and dependence; they do not rank “power” or predict crises on their own.

Definition. Import substitution industrialization is a strategy where a country tries to cut dependence on imports by building domestic industries. Tariffs, quotas, or directed credit often support that shift.

Key idea. Protection, domestic production, and industrial policy steer demand toward local suppliers instead of foreign goods—so you look for industrial shares moving while import intensity changes.

How to detect it in data

• Imports (% of GDP) flatten or fall—NE.IMP.GNFS.ZS in the ISI diagnostics study.
• Manufacturing share of GDP rises—NV.IND.MANF.ZS in the same panel.
• Broader industry share rises—NV.IND.TOTL.ZS (manufacturing sits inside industry in the accounts).
• Exports stay weak or lag imports—compare NE.EXP.GNFS.ZS with NE.IMP.GNFS.ZS on the trade-structure chart.
• Growth may later slow if competitiveness or productivity stalls—NY.GDP.MKTP.KD.ZG is only a coarse outcome read, not a productivity measure.

What to look for in charts

• Imports stop rising or trend down while manufacturing (and often industry) picks up early.
• Exports lag: the export side of the trade panel does not keep pace with the import side you are trying to replace.
• Indexed overview (base year = 100): good for timing and co-movement of the four structure series—not for reading absolute % of GDP levels.
• Later windows: stagnation or partial reversal in industrial lines, or import pressure returning—descriptive timing only.

Related study. ISI diagnostics — trade and industry structure

Definition. Dutch disease is the idea that a resource boom—often in oil—shifts the economy in ways that weaken other tradable sectors. Manufacturing is the usual sector people track.

Key idea. Resource revenues can strengthen the real exchange rate and pull spending and labor; manufacturing and other tradables then face tougher competition at home and abroad.

How to detect it in data

• Oil or resource rents rise as % of GDP—NY.GDP.PETR.RT.ZS on the Dutch disease diagnostics study.
• Manufacturing share of GDP falls—NV.IND.MANF.ZS on the same study.
• Imports rise as % of GDP—NE.IMP.GNFS.ZS (import intensity climbing while tradables struggle).
• Open-market USD→toman in that study adds FX pressure next to the annual WDI lines.
• Consumption rising and investment in tradables weakening are visible in Iran national-accounts studies (composition / dual-axis)—use them alongside this panel, which does not plot C and I.

What to look for in charts

• Resource dependence (rents) trends up while manufacturing trends down—the core tension to watch.
• Imports climb in parallel with strong resource income—suggests spending and exchange-rate pressure; crude can move for other reasons too.
• The pattern is a persistent structural gap over years, not a single-year blip—use event overlays as timing anchors, not automatic proof of cause.

Related study. Dutch disease diagnostics — Iran (pattern view)

Many economies report an official or policy-influenced rate while people trade at a market rate. SignalMap plots both where data allow so you can see the wedge—not to pick which rate is “true,” but to show that multiple prices can coexist.

A spread is a derived percentage gap between two rates, e.g. (open − official) ÷ official. It summarizes how wide the wedge is at a point in time; it is descriptive and sensitive to definitions of each series.

Capital controls limit money moving across borders; they often accompany gaps between official and market FX. Parallel or informal markets reflect what people pay when formal channels are tight—measurements rarely capture every transaction.

High inflation erodes domestic purchasing power; the exchange rate shows how local currency trades against others. Studies use CPI, FX, and spreads as separate lenses—each has limits and none alone explains the whole economy.

When two series use different y-axes (e.g. oil on the left, FX on the right), the chart shows timing and co-movement. You should not compare the left numbers directly to the right numbers or infer one caused the other without separate evidence.

Indexing sets a base period to 100 so series with different units can be compared for relative change. Good for “Iran vs Turkey” burden charts; you lose absolute levels on the same scale—read the study notes for what the base year means.

A log y-axis makes equal distances mean similar percentage moves. It helps when early years are tiny next to recent values (e.g. long-range gold or wide-ranging prices). It does not change the underlying data.

Some series are annual (PPP, many WDI panels); others are daily (spot FX or Brent). Overlaying them is for broad timing context—do not read day-to-day wiggles as if they were measured at the same frequency as annual points.

Vertical or shaded markers place events in time—wars, sanctions, political milestones. They are anchors for reading charts, not proof that an event moved the series.

SignalMap shows patterns and context. It does not assert causality, forecast the future, or claim that any marker explains a change in the data by itself.

Sanctions limit trade, finance, or specific activities. They show up in studies as context layers; they are not modeled as automatic impacts on every series.

When shipments are constrained, the bottleneck is often how much crude can leave the country, not the headline world price. That is why capacity-style charts pair price with volume estimates.

When official channels are tight, transactions move outside formal measurement. Open-market rates and trade estimates can reflect that reality only partially.

A supply shock changes how much can be produced or sold (war, outage, OPEC decision). A demand shock changes how much buyers want. Prices can move for either reason; overlays do not say which mechanism dominated.

Spreads, real or indexed values, PPP burdens, and capacity proxies are computed from base data. Definitions and bases are given in each study’s sources; small changes in inputs or windows change the derived line.

Two series moving together are correlated. That does not mean one caused the other. Causation needs a separate argument and usually much more than a time series.

Published indicators are revised, miss informal activity, and use definitions that may not match everyday experience. Treat numbers as useful approximations, not complete pictures.

Linear growth adds a roughly fixed amount per time step. Exponential growth multiplies by a factor each step—early values look flat until the curve steepens. Follower-count demos use both for contrast.

Logistic growth rises quickly then levels off as it approaches a ceiling—useful when a quantity cannot grow without limit (e.g. saturation of an audience).

Fitting means choosing parameters so a curve sits close to past points. Flexible curves can hug noise (overfitting) and look worse on new data; simpler curves are often more honest for illustration.

Discourse studies treat comment text as data: words are counted, weighted, and grouped to summarize themes. Results depend on language, sampling, and preprocessing.

TF-IDF highlights words that are frequent in one document (or comment) but uncommon in the whole collection—useful for picking distinctive terms without hand-picking every keyword.

High-dimensional vectors (e.g. from text) are projected to 2D for visualization. PCA stresses global spread; UMAP stresses local neighborhoods and clusters. Both are exploratory, not ground truth about “topics.”

This tool labels transcript chunks with candidate rhetorical patterns (e.g. types of potential fallacies). It is experimental: labels are aids for exploration, not proof that a formal fallacy occurred. Three method families are available for comparison: rule-based heuristics, a future classifier model, and an LLM-assisted pass. Each method uses different signals, so they will often disagree — that disagreement is informative, not a bug.

Uses hand-tuned keyword and phrase detectors with simple context guards. What it uses: explicit string patterns and lightweight rules over chunk text. Strengths: transparent, reproducible, and cheap to run; good for baseline coverage and debugging. Weaknesses: English-centric, brittle to paraphrase, and cannot capture full argument structure. Language support: English only for now; Persian heuristics are scaffolded in the codebase but not executed, so Persian transcripts get analysis_supported=false with an explicit note. When it may fail: sarcasm, code-switching, implicit premises, or valid rhetoric that resembles a pattern. It may disagree with the LLM because the LLM infers intent and paraphrase while heuristics only match surface cues.

Will use a supervised or embedding-based model trained on labeled examples (or similar signals) rather than fixed keyword lists. What it uses: learned weights from data — typically text features or dense embeddings — mapped to fallacy categories. Strengths: can generalize beyond exact phrases if training matches the domain. Weaknesses: depends on label quality, class balance, and domain shift; errors can be opaque without careful evaluation. When it may fail: out-of-domain topics, rare phrasing, or labels that do not match how annotators defined fallacies. Not yet implemented in this app; results are disabled until the pipeline is ready.

Uses a hosted large language model with structured JSON prompts to assign labels and short rationales per chunk. What it uses: semantic reasoning over the transcript text via the model’s weights (not audio diarization). Strengths: handles varied wording and can supply explanations. Language support: for pasted transcripts, pick English or Farsi in the language control; the model uses dedicated English and Persian system prompts (same JSON schema). YouTube captions supply language automatically. Other languages may use the English prompt with an API note. Weaknesses: non-deterministic across runs, possible hallucinations, and sensitivity to prompt wording; not a substitute for domain validation. When it may fail: subtle logic, unstated assumptions, or chunks where the model overfits to keywords. It may disagree with heuristics because it interprets meaning more freely, or with a future classifier because training objectives differ.

Heuristics surface explicit cues; classifiers optimize for training distributions; LLMs infer loosely from natural language. The same chunk might trigger a rule, score high in a model, and be rejected by an LLM — or the reverse. Treat disagreement as a signal to read the source text, not as proof that one method is “right.” Compare outputs cautiously and avoid using any single method alone for high-stakes conclusions.