fix minor errors

app.py

@@ -4,6 +4,7 @@ from fixed_recall import FixedRecall
 import evaluate
 import gradio as gr
 import pandas as pd
+import numpy as np
 
 title = "'Combine' multiple metrics with this 🤗 Evaluate 🪲 Fix!"
 
@@ -14,10 +15,29 @@ Check out the original, longstanding issue [here](https://github.com/huggingface
 `evaluate.combine()` multiple metrics related to multilabel text classification. Particularly, one cannot `combine` the `f1`, `precision`, and `recall` scores for \
 evaluation. I encountered this issue specifically while training [RoBERTa-base-DReiFT](https://huggingface.co/MarioBarbeque/RoBERTa-base-DReiFT) for multilabel \
 text classification of 805 labeled medical conditions based on drug reviews. The [following workaround](https://github.com/johngrahamreynolds/FixedMetricsForHF) was 
-created to address this. \n
+created to address this - follow the link to view the source! \n
 
 This Space shows how one can instantiate these custom `evaluate.Metric`s, each with their own unique methodology for averaging across labels, before `combine`-ing them into a  
-HF `evaluate.CombinedEvaluations` object. From here, we can easily compute each of the metrics simultaneously using `compute`.</p>
+HF `evaluate.CombinedEvaluations` object. From here, we can easily compute each of the metrics simultaneously using `compute`. \n
+
+In general, one writes the following:\n
+
+```python
+f1 = FixedF1(average=...)
+precision = FixedPrecision(average=...)
+recall = FixedRecall(average=...)
+
+combined = evaluate.combine([f1, precision, recall])
+
+combined.add_batch(predictions=..., references=...)
+combined.compute()
+```\n
+
+where the `average` parameter can be different at instantiation time for each of the metrics. Acceptable values include `[None, 'micro', 'macro', 'weighted']` (
+or `binary` if there exist only two labels). \n
+
+Try it out using the examples below! Then try picking some various averaging methods yourself!
+</p>
 """
 
 
@@ -39,13 +59,13 @@ def evaluation(predictions_df: pd.DataFrame, metrics_df: pd.DataFrame) -> str:
     combined_list = []
 
     if "f1" in metric_set:
-        f1 = FixedF1(average=metric_map["f1"])
+        f1 = FixedF1(average=metric_map["f1"] if metric_map["f1"] != "None" else None)
         combined_list.append(f1)
     if "precision" in metric_set:
-        precision = FixedPrecision(average=metric_map["f1"])
+        precision = FixedPrecision(average=metric_map["precision"] if metric_map["precision"] != "None" else None, zero_division=np.nan)
         combined_list.append(precision)
     if "recall" in metric_set:
-        recall = FixedRecall(average=metric_map["f1"])
+        recall = FixedRecall(average=metric_map["recall"] if metric_map["recall"] != "None" else None)
         combined_list.append(recall)
     
     combined = evaluate.combine(combined_list)