initial commit of notebook exploring child mortality data

Author: Elye Bliss

notebooks/2025_09_16_child_mortality_investigation.py

@@ -0,0 +1,100 @@
+import seaborn as sns
+import plotly.express as px
+import pandas as pd
+import matplotlib.pyplot as plt
+import matplotlib.ticker as mticker
+
+
+DATA_PATH = "/mnt/team/rapidresponse/pub/population/modeling/climate_malnutrition/child_mortality/training_data/2025_09_15.01/data.parquet"
+
+df = pd.read_parquet(DATA_PATH)
+
+df["line_id"] = df["line_id"].astype(int)
+
+df["indv_id"] = df[["nid", "psu", "hh_id", "line_id"]].astype(str).agg("_".join, axis=1)
+print(f"{df['indv_id'].nunique():,} unique individuals in data")
+
+# flip child_alive so 1 = died, 0 = alive for easier interpretation
+df["child_mortality"] = 1 - df["child_alive"]
+
+# get unique individuals per year
+agg_yr = (
+    df.groupby(["int_year"])["indv_id"]
+    .nunique()
+    .reset_index()
+    .rename(columns={"indv_id": "unique_individuals"})
+)
+plt.figure(figsize=(20, 5))
+ax = agg_yr.plot(x="int_year", y="unique_individuals", kind="bar")
+plt.title("Unique Individuals per Year")
+plt.xticks(fontsize=5, rotation=45, ha="right")
+ax.yaxis.set_major_formatter(mticker.FuncFormatter(lambda x, p: format(int(x), ",")))
+plt.tight_layout()
+plt.show()
+
+# get max age of children who did not die
+children_died_ids = df.query("child_alive == 0")["indv_id"].unique()
+children_alive = df.query("indv_id not in @children_died_ids")
+children_alive["age_month_original"] = children_alive["age_month_original"].astype(int)
+children_alive["over_5_at_survey"] = children_alive["age_month_original"] > 60
+over5_at_survey = children_alive[children_alive["over_5_at_survey"]][
+    "indv_id"
+].nunique()
+under5_at_survey = children_alive[~children_alive["over_5_at_survey"]][
+    "indv_id"
+].nunique()
+print(
+    f"{under5_at_survey:,} children out of {over5_at_survey + under5_at_survey:,} "
+    f"who did not die were under 5 at survey ({under5_at_survey/(over5_at_survey + under5_at_survey):.1%})"
+)
+
+# Aggregate data
+agg_df = (
+    df.groupby(["nid", "ihme_loc_id", "int_year"], as_index=False)
+    .mean(numeric_only=True)
+    .rename(
+        columns={
+            "ldipc_weighted_no_match": "consumption",
+            "child_mortality": "total_mortality",
+        }
+    )
+)
+px.scatter(agg_df, x="mean_temperature", y="total_mortality", color="ihme_loc_id")
+
+
+columns_to_bin = [
+    "mean_temperature",
+    "total_precipitation",
+    "relative_humidity",
+    "mean_high_temperature",
+    "mean_low_temperature",
+    "precipitation_days",
+    "days_over_30C",
+    "days_over_26C",
+]
+heatmap_df = df.copy()  # .query("ihme_loc_id == 'ETH'")
+for col in columns_to_bin:
+    heatmap_df[f"{col}_bin"] = pd.qcut(
+        heatmap_df[col], 10, retbins=False, duplicates="drop"
+    )
+heatmap_df["consumption"], ldi_bins = pd.qcut(
+    heatmap_df.ldipc_weighted_no_match, 10, retbins=True
+)
+col = "days_over_30C"
+sns.heatmap(
+    heatmap_df.groupby(["consumption", f"{col}_bin"])["child_mortality"]
+    .mean()
+    .unstack(),
+    annot=True,
+    fmt=".2f",
+    cmap="YlOrBr",
+)
+col = "mean_temperature"
+sns.heatmap(
+    heatmap_df.groupby(["consumption", f"{col}_bin"])["child_mortality"]
+    .mean()
+    .unstack(),
+    annot=True,
+    fmt=".2f",
+    cmap="YlOrBr",
+)