import zipfile
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

sns.set_theme(style="whitegrid", context="talk")
pd.set_option("display.max_columns", 40)

ZIP = "/srv/data/giatar-invasive/GIATAR - Data and query code.zip"
BASE = "GIATAR - Data and query code/dataset/"

z = zipfile.ZipFile(ZIP)
# Show the key data files (skip folders and query code)
for info in z.infolist():
    name = info.filename
    if name.startswith(BASE) and name.endswith(".csv") and (
        "occurrences/" in name or "species lists/invasive_all_source" in name):
        print(f"{info.file_size/1e6:6.1f} MB   {name[len(BASE):]}")

  96.7 MB   occurrences/all_records.csv
  64.8 MB   occurrences/first_records.csv
  11.0 MB   species lists/invasive_all_source.csv

with z.open(BASE + "occurrences/first_records.csv") as f:
    fr = pd.read_csv(f, usecols=["usageKey", "ISO3", "year", "Source", "Type"],
                     dtype={"usageKey": "string", "ISO3": "string"})

print("raw shape:", fr.shape)

# Clean the year: drop missing, keep a plausible modern window (1500-2025)
fr = fr.dropna(subset=["year", "ISO3"])
fr["year"] = fr["year"].astype(int)
before = len(fr)
fr = fr[(fr["year"] >= 1500) & (fr["year"] <= 2025)]
print(f"kept {len(fr):,} of {before:,} rows after year cleaning "
      f"({len(fr)/before:.1%})")
print(f"taxa: {fr['usageKey'].nunique():,}   countries (ISO3): {fr['ISO3'].nunique()}")
fr.head()

raw shape: (898007, 5)

kept 826,492 of 827,300 rows after year cleaning (99.9%)
taxa: 44,816   countries (ISO3): 274

per_year = fr.groupby("year").size()
# Restrict the visual to the well-sampled modern era
per_year = per_year[per_year.index >= 1800]
decade = per_year.groupby((per_year.index // 10) * 10).sum()
cumulative = per_year.sort_index().cumsum()

fig, ax1 = plt.subplots(figsize=(12, 6))
ax1.bar(decade.index, decade.values, width=8, color="#4C72B0",
        alpha=0.85, label="New first records per decade")
ax1.set_xlabel("Year")
ax1.set_ylabel("New first records per decade", color="#4C72B0")
ax1.tick_params(axis="y", labelcolor="#4C72B0")

ax2 = ax1.twinx()
ax2.plot(cumulative.index, cumulative.values, color="#C44E52", lw=3,
         label="Cumulative")
ax2.set_ylabel("Cumulative first records", color="#C44E52")
ax2.tick_params(axis="y", labelcolor="#C44E52")
ax2.grid(False)

ax1.set_title("Accumulation of invasive first records worldwide (1800-2025)")
fig.tight_layout()
plt.show()

# Distinct taxa per country
taxa_per_country = (fr.groupby("ISO3")["usageKey"].nunique()
                      .sort_values(ascending=False).head(20))

# Taxonomy lookup from the species list (one row per usageKey)
with z.open(BASE + "species lists/invasive_all_source.csv") as f:
    sp = pd.read_csv(f, usecols=["usageKey", "canonicalName", "kingdom"],
                     dtype={"usageKey": "string"})
sp = sp.dropna(subset=["usageKey"]).drop_duplicates("usageKey")

taxa = fr[["usageKey"]].drop_duplicates().merge(sp, on="usageKey", how="left")
kingdom = taxa["kingdom"].fillna("Unclassified").value_counts().head(8)

fig, axes = plt.subplots(1, 2, figsize=(16, 7))
sns.barplot(x=taxa_per_country.values, y=taxa_per_country.index,
            color="#55A868", ax=axes[0])
axes[0].set_title("Top 20 countries by number of\ninvasive/alien taxa recorded")
axes[0].set_xlabel("Distinct taxa"); axes[0].set_ylabel("Country (ISO3)")

sns.barplot(x=kingdom.values, y=kingdom.index, color="#8172B3", ax=axes[1])
axes[1].set_title("Invasive/alien taxa by kingdom")
axes[1].set_xlabel("Number of taxa"); axes[1].set_ylabel("")
fig.tight_layout()
plt.show()

# Most widespread taxon in the cleaned table
widest = fr.groupby("usageKey")["ISO3"].nunique().sort_values(ascending=False)
focus_key = widest.index[0]
focus_name = sp.set_index("usageKey")["canonicalName"].get(focus_key, focus_key)

sub = (fr[fr["usageKey"] == focus_key]
       .sort_values("year")
       .drop_duplicates("ISO3", keep="first"))
spread = sub.groupby("year").size().sort_index().cumsum()

fig, ax = plt.subplots(figsize=(12, 6))
ax.step(spread.index, spread.values, where="post", color="#C44E52", lw=3)
ax.fill_between(spread.index, spread.values, step="post", alpha=0.15, color="#C44E52")
ax.set_title(f"Global spread of {focus_name}\n(cumulative countries with a first record)")
ax.set_xlabel("Year of first record")
ax.set_ylabel("Countries reached (cumulative)")
ax.margins(x=0.01)
print(f"{focus_name}: recorded in {sub['ISO3'].nunique()} countries; "
      f"earliest {int(sub['year'].min())}, latest {int(sub['year'].max())}")
plt.tight_layout()
plt.show()

Animalia: recorded in 252 countries; earliest 1963, latest 2019

# ⚠️ RESTORE: this DISCARDS YOUR CHANGES to this notebook and resets it to the original.
# 1. Uncomment the line below (remove the #)   2. Run this cell
# 3. Then: menu File → Reload Notebook from Disk

# !git -C ~/citizen-science-data fetch -q origin && git -C ~/citizen-science-data checkout origin/main -- giatar-invasive.ipynb && echo "Restored. Now: File → Reload Notebook from Disk"

	usageKey	ISO3	year	Source	Type
0	1	ABW	1970	GBIF	First report
1	1	AFG	1970	GBIF	First report
2	1	AGO	1970	GBIF	First report
3	1	AIA	1971	GBIF	First report
4	1	ALA	1970	GBIF	First report

GIATAR: Global Invasive and Alien Traits and Records Dataset¶

What's in the archive¶

The first-records table¶

1. The invasion accumulation curve¶

2. Invasion hotspots and who the invaders are¶

3. Tracking the spread of one taxon¶

Takeaways¶

Your turn¶