메모리 최적화

baseline = df.memory_usage(deep=True).sum() / 1024**2
print(f"baseline: {baseline:.2f} MB")

int_cols = df.select_dtypes(include=["int64"]).columns
float_cols = df.select_dtypes(include=["float64"]).columns

df[int_cols] = df[int_cols].apply(pd.to_numeric, downcast="integer")
df[float_cols] = df[float_cols].apply(pd.to_numeric, downcast="float")

for c in ["region", "device", "segment"]:
    df[c] = df[c].astype("category")

chunks = pd.read_csv("events.csv", chunksize=200_000)
agg = []
for chunk in chunks:
    part = chunk.groupby("event_type")["amount"].sum()
    agg.append(part)
result = pd.concat(agg).groupby(level=0).sum()