msminhas93

for row in document.select(&row_selector) {
    let title_selector = Selector::parse("td.title > span.titleline > a")?;
    if let Some(title_element) = row.select(&title_selector).next() {
        let title = title_element.inner_html();
        let link = title_element.value().attr("href").unwrap_or("").to_string();
        submissions.push(Submission { title, link });
    }
}

msminhas93

let document = Html::parse_document(&response);
let row_selector = Selector::parse("tr.athing")?;

msminhas93

let url = format!("<https://news.ycombinator.com/news?p={}>", page);
let response = get(&url)?.text()?;

msminhas93

pub fn fetch_hacker_news(page: usize) -> Result<Page, Box<dyn Error>> {
    // ... (implementation details)
}

msminhas93

pub struct Submission {
    pub title: String,
    pub link: String,
}

pub struct Page {
    pub submissions: Vec<Submission>,
    pub current_page: usize,
    pub total_pages: usize,
}

msminhas93

# Your code here
print("Hello, World!")

msminhas93

Metric,Behavior Across Classifier Types,Impact of Class Imbalance,When to Use,Key Considerations
Accuracy,"- Random: Equal to 0.5. - All 1: Equals positive class proportion - All 0: Equals negative class proportion","- Highly sensitive; can be misleading in imbalanced datasets","- Use in balanced datasets or when all classes are equally important","- Can be deceptive in imbalanced datasets; consider other metrics for a more comprehensive evaluation"
Precision,"- Random: Macro average is equal to 0.5. - All 1: Class 0 precision is 0. - All 0: Class 1 precision 0.","- Cases where the number of actual positives is very, very low, say 1-2 examples in total, precision is less meaningful and less useful as a metric","- When the cost of false positives is high (e.g., spam detection, medical diagnosis)","- Should be balanced with recall; precision alone may not capture overall performance"
Recall,"- Random: Equal to 0.5 - All 1: Macro average 0.5 always. - All 0: Macro average always 0.5","- In an imbalanced dataset 

msminhas93

# if you want to iterate over all the samples in the current view
for sample in session.view:
    softmax_values = torch.softmax(torch.from_numpy(sample.predictions.logits),0).numpy()
    wine_confidence = softmax_values[red_wine_index] + softmax_values[wine_glass_index]
    bagel_confidence = softmax_values[bagel_index]
    print(wine_confidence, bagel_confidence)
# if you want to iterate over the selected samples in the datset
for sample in dataset[session.selected]:
    softmax_values = torch.softmax(torch.from_numpy(sample.predictions.logits),0).numpy()
    wine_confidence = softmax_values[red_wine_index] + softmax_values[wine_glass_index]

msminhas93

cm = results.plot_confusion_matrix()
cm.show()
session.plots.attach(cm)

msminhas93

results.print_report()