Add pred acc and ROC curve

pf2/figures/figureA9.py

@@ -0,0 +1,108 @@
+"""
+Figure A8:
+"""
+import numpy as np
+import pandas as pd
+import anndata 
+from sklearn.metrics import accuracy_score
+import seaborn as sns
+from ..data_import import convert_to_patients, import_meta
+from ..predict import predict_mortality
+from .common import subplotLabel, getSetup
+from sklearn.metrics import RocCurveDisplay
+
+
+def makeFigure():
+    """Get a list of the axis objects and create a figure."""
+    ax, f = getSetup((6, 3), (1, 2))
+    subplotLabel(ax)
+
+    X = anndata.read_h5ad("/opt/northwest_bal/full_fitted.h5ad")
+
+    meta = import_meta()
+    conversions = convert_to_patients(X)
+
+    patient_factor = pd.DataFrame(
+        X.uns["Pf2_A"],
+        index=conversions,
+        columns=np.arange(X.uns["Pf2_A"].shape[1]) + 1,
+    )
+    meta = meta.loc[patient_factor.index, :]
+
+    plsr_acc_df = plsr_acc(patient_factor, meta)
+
+    sns.barplot(data=plsr_acc_df, ax=ax[0])
+    ax[0].set(ylim=[0, 1], ylabel="Accuracy")
+
+    plot_plsr_auc_roc(patient_factor, meta, ax[1])
+
+    return f
+
+
+def plsr_acc(patient_factor_matrix, meta_data):
+    """Runs PLSR and obtains average prediction accuracy"""
+
+    acc_df = pd.DataFrame(columns=["Overall", "C19", "nC19"])
+
+    probabilities, labels = predict_mortality(
+        patient_factor_matrix,
+        meta_data,
+        proba=True
+    )
+
+    probabilities = probabilities.round().astype(int)
+    meta_data = meta_data.loc[~meta_data.index.duplicated()].loc[labels.index]
+
+    covid_acc = accuracy_score(
+        labels.loc[meta_data.loc[:, "patient_category"] == "COVID-19"],
+        probabilities.loc[meta_data.loc[:, "patient_category"] == "COVID-19"]
+    )
+    nc_acc = accuracy_score(
+        labels.loc[meta_data.loc[:, "patient_category"] != "COVID-19"],
+        probabilities.loc[meta_data.loc[:, "patient_category"] != "COVID-19"]
+    )
+    acc = accuracy_score(labels, probabilities)
+
+    acc_df.loc[
+        0,
+        :
+        ] = [acc, covid_acc, nc_acc]
+
+    return acc_df
+
+
+def plot_plsr_auc_roc(patient_factor_matrix, meta_data, ax):
+    """Runs PLSR and plots ROC AUC based on actual and prediction labels"""
+
+    probabilities, labels = predict_mortality(
+        patient_factor_matrix,
+        meta_data,
+        proba=True
+    )
+
+    probabilities = probabilities.round().astype(int)
+    meta_data = meta_data.loc[~meta_data.index.duplicated()].loc[labels.index]
+
+    # covid_acc = accuracy_score(
+    #     labels.loc[meta_data.loc[:, "patient_category"] == "COVID-19"],
+    #     probabilities.loc[meta_data.loc[:, "patient_category"] == "COVID-19"]
+    # )
+    # nc_acc = accuracy_score(
+    #     labels.loc[meta_data.loc[:, "patient_category"] != "COVID-19"],
+    #     probabilities.loc[meta_data.loc[:, "patient_category"] != "COVID-19"]
+    # )
+    # acc = 
+    # # c19, nc19 = predict_mortality(
+    # #     patient_factor_matrix,
+    # #     meta_data,
+    # #     auc_roc=True
+    # # )
+    # print(c19[0])
+    # print(c19[1])
+    RocCurveDisplay.from_predictions(labels.loc[meta_data.loc[:, "patient_category"] == "COVID-19"], 
+                                     probabilities.loc[meta_data.loc[:, "patient_category"] == "COVID-19"],
+                                     ax=ax, name="C19")
+    RocCurveDisplay.from_predictions(labels.loc[meta_data.loc[:, "patient_category"] != "COVID-19"],
+                                     probabilities.loc[meta_data.loc[:, "patient_category"] != "COVID-19"], 
+                                     ax=ax, name="nC19")
+

pf2/predict.py

@@ -47,12 +47,14 @@ def run_plsr(
         plsr.coef_.squeeze(),
         index=data.columns
     )
-
+    
     if proba:
         return probabilities, plsr
+
     else:
         predicted = probabilities.round().astype(int)
         return predicted, plsr
+
 
 
 def predict_mortality(
@@ -101,7 +103,7 @@ def predict_mortality(
         meta.loc[:, "patient_category"] != "COVID-19",
         "binary_outcome"
     ]
-
+    
     predictions = pd.Series(index=data.index)
     predictions.loc[meta.loc[:, "patient_category"] == "COVID-19"], c_plsr = \
         run_plsr(
@@ -114,6 +116,7 @@ def predict_mortality(
 
     if proba:
         return predictions, labels
+
     else:
         predicted = predictions.round().astype(int)
         return accuracy_score(labels, predicted), (c_plsr, nc_plsr)