cleaner.py

import csv
import json

# Define the header names so we can specify them when reading the CSV
fieldnames = (
    "STATION","NAME","DATE","AWND","PRCP","SNOW","SNWD",
    "TAVG","TMAX","TMIN","WT01","WT02","WT03","WT05","WT08"
)

# Open both files so we can interact with them
# Using the `with` keyword lets us close the files automatically after these
# with blocks end and we're done writing and reading the files
with open('./seattle-data.csv', 'r') as csvfile:
    with open('./seattle-data.json', 'w') as jsonfile:
        # `next` will simply skip over the header row in the csvfile
        next(csvfile)
        # We use the csv library to create a 'reader' of the file
        # This reader patse through the csvfile and the headers 
        # and allow us to interact with it as a Python object
        reader = csv.DictReader(csvfile, fieldnames)
        # This creates an empty dictionary to hold the final set of 
        # data that we'll eventually dump into the JSON file
        final_data = {}
        # Now we use the reader to iterate over all the rows of the CSV
        # (except for the header) and then keep the values we want 
        for row in reader:
            # We also restructure the data so that it exists as 
            # a set of date keys with the value as a dictionary of
            # different data elements from the CSV.
            final_data[row["DATE"]] = {
                "AWND": row["AWND"],
                "PRCP": row["PRCP"],
                "SNOW": row["SNOW"],
                "SNWD": row["SNWD"],
                "TAVG": row["TAVG"],
                "TMAX": row["TMAX"],
                "TMIN": row["TMIN"]
            }
        # Finally, we use the json library to output the final_data
        # dictionary to the jsonfile we opened earlier
        json.dump(final_data, jsonfile)
        # And then we write a final newline to the end of the file 
        # (this is just a best practice)
        jsonfile.write('\n')