import re
import statistics

# Get the markdown table from the input field
markdown_table = pd['steps']['trigger']['data']['markdown_table']

ratings = []
types_of_result = []

# Match the type of result and information gain rating using a regular expression
pattern = re.compile(r'\|\s*(.*?)\s*\|\s*(.+?)\s*\|\s*(\d+)\s*\|\s*(.+?)\s*\|')

# Find all matches and add the rating and type of result to the lists
for match in pattern.findall(markdown_table):
    types_of_result.append(match[0])
    ratings.append(int(match[2]))

# Calculate the average of the ratings and round to the nearest integer
if ratings:
    avg_rating = round(statistics.mean(ratings))
else:
    avg_rating = 0

# Count the number of rows in the table (excluding headings)
num_rows = markdown_table.count('\n') - 3

# Determine the most common type(s) of result
if types_of_result:
    freq_dict = {}
    for type in types_of_result:
        if type in freq_dict:
            freq_dict[type] += 1
        else:
            freq_dict[type] = 1
    max_freq = max(freq_dict.values())
    freq_types = [type for type, freq in freq_dict.items() if freq == max_freq]
    median_type = ','.join(freq_types)
else:
    median_type = ''

# Define the output dictionary
output_data = {
    'ratings': ','.join(map(str, ratings)),
    'average_rating': avg_rating,
    'num_rows': num_rows,
    'median_type': median_type
}

# Return the output dictionary
return output_data