# Sample Queries

All queries assume you are in the `historysaid-global-economic-dataset/` directory.

## Get GDP for a specific country

### Python
```python
import pandas as pd
df = pd.read_parquet("data/unified/all_indicators.parquet")
turkey_gdp = df[(df["country_code"] == "TUR") & (df["indicator_id"] == "wb.NY.GDP.MKTP.CD")]
turkey_gdp = turkey_gdp.dropna(subset=["value"]).sort_values("year")
print(turkey_gdp[["year", "value"]].tail(10))
```

### R
```r
library(readr); library(dplyr)
df <- read_csv("data/unified/all_indicators.csv", show_col_types = FALSE)
df %>% filter(country_code == "TUR", indicator_id == "wb.NY.GDP.MKTP.CD", !is.na(value)) %>%
  arrange(year) %>% select(year, value) %>% tail(10) %>% print()
```

## Compare an indicator across countries

### Python
```python
countries = ["USA", "CHN", "IND", "BRA", "NGA"]
inflation = df[(df["indicator_id"] == "wb.FP.CPI.TOTL.ZG") & (df["country_code"].isin(countries)) & (df["year"] == 2023)]
print(inflation[["country_name", "year", "value"]].sort_values("value"))
```

### R
```r
df %>% filter(indicator_id == "wb.FP.CPI.TOTL.ZG",
              country_code %in% c("USA", "CHN", "IND", "BRA", "NGA"),
              year == 2023) %>%
  select(country_name, year, value) %>% arrange(value) %>% print()
```

## Find all available indicators for a country

### Python
```python
japan = df[df["country_code"] == "JPN"]
indicators = japan.groupby(["indicator_id", "indicator_name"]).agg(
    years=("year", "count"),
    non_null=("value", "count")
).reset_index()
print(indicators.sort_values("indicator_id"))
```

### R
```r
df %>% filter(country_code == "JPN") %>%
  group_by(indicator_id, indicator_name) %>%
  summarise(years = n(), non_null = sum(!is.na(value)), .groups = "drop") %>%
  arrange(indicator_id) %>% print(n = Inf)
```

## Plot a time series

### Python
```python
import matplotlib.pyplot as plt
pop = df[(df["country_code"] == "DEU") & (df["indicator_id"] == "wb.SP.POP.TOTL")]
pop = pop.dropna(subset=["value"]).sort_values("year")
plt.plot(pop["year"], pop["value"] / 1e6)
plt.title("Germany Population (millions)")
plt.xlabel("Year")
plt.ylabel("Millions")
plt.grid(True)
plt.tight_layout()
plt.show()
```

### R
```r
library(ggplot2)
df %>% filter(country_code == "DEU", indicator_id == "wb.SP.POP.TOTL", !is.na(value)) %>%
  ggplot(aes(x = year, y = value / 1e6)) +
  geom_line() +
  labs(title = "Germany Population (millions)", x = "Year", y = "Millions") +
  theme_minimal()
```

## Check data coverage for an indicator

### Python
```python
import json
with open("_mappings/coverage_matrix.json") as f:
    coverage = json.load(f)
for c in sorted(coverage, key=lambda x: x["fill_rate_pct"], reverse=True)[:10]:
    print(f"{c['indicator_id']}: {c['countries_with_data']} countries, {c['fill_rate_pct']}% fill rate")
```

### R
```r
library(jsonlite)
coverage <- fromJSON("_mappings/coverage_matrix.json")
coverage %>% arrange(desc(fill_rate_pct)) %>% head(10) %>%
  select(indicator_id, countries_with_data, fill_rate_pct) %>% print()
```