Python Forum

I have most of the code written. I want to add the first two columns of the csv file to be date and time. And, have DATE and TIME in the headers. When I comment out the date and time, it exports everything uniformly, just under the wrong columns.

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Wed Nov  2 23:17:11 2022

@author: eh5713
"""
#script to communicate with arduino and load serial monitor data
#Your sketch/code should be already loaded/flashed to the board prior to running this script
#However communication port should not be occupied
#I usually close Arduino software after flashing the code to be on the safe side

#we need serial module to communicate with the arduino
#csv module is needed because we will be logging sensor data to a csv file

import serial
import csv
from datetime import date,datetime


arduino_port="COM3" #serial port of arduino
#Port that arduino uses - for my mac it is usbmodem.., on windows it will be COM3-4
baud = 9600
#defining the file we will use for data loading

#creates a file named by the user's input
filename = input("Input the Filename: ")
f_extns = filename.split(".csv")
print ("The extension of the file is : " + repr(f_extns[-1]))


#filename = "data_loaded.csv"

# open serial port
ser = serial.Serial(arduino_port, baud)

#print out the active port
print("Connected to Arduino port:" + arduino_port)
#we create 'data_loaded.csv file'
file = open(filename,'w')
print("File created!")

#creating empty lists for data to be taken



#this is a loop that performs 20 consecutive readings
samples = 20
print_labels = False
line = 0
sensor_data = []
#dates = []
#instance = []

while line <= samples:
    getData=ser.readline()
    dataString = getData.decode('utf-8')
    data = dataString[0:][:-2]
    print(data)

    # datetime object containing current date and time
    #data_taken = datetime.now()
    #data_taken_date=data_taken.strftime('%m/%d/%Y')
    #dates.append(data_taken_date)
    #data_taken_time=data_taken.strftime('%H:%M:%S')
    #instance.append(data_taken_time)

    readings = data.split(",")
    print(readings)
    sensor_data.append(readings)
    print(sensor_data)
    #overall=list(zip(dates,instance,sensor_data))

    line = line + 1
with open(filename, 'w',encoding = 'UTF8', newline= '') as f:
    writer = csv.writer(f)
    writer.writerows(date,sensor_data)
       
#closing the port
file.close()
ser.close()

The reason this didn't work:

overall=list(zip(dates,instance,sensor_data))

Is that sensor_data is a list of lists so you get date, time, [sensor, data, as, a, list]. You need to flatten the sensor data. This is most easily done while the data is collected.

import csv
from datetime import datetime

class Serial:
    """My fake serial port listening to my imaginary Arduino"""
    def __init__(self, *args, **kwargs):
        print("opening", args, kwargs)

    def close(self):
        print("closing")

    def __enter__(self, *args, **kwargs):
        return self

    def __exit__(self, *args, **kwargs):
        self.close()

    def readline(self):
        return b"1,2,3,4,5\r\n"

labels = ["date", "time", "s1", "s2", "s3", "s4", "s5"]
samples = 20
rows = []

with Serial("COM3", 9600) as ser:
    for _ in range(samples):
        data = ser.readline()[:-2].decode('utf-8').split(",")
        now = datetime.now()
        rows.append([
            now.strftime('%m/%d/%Y'),
            now.strftime('%H:%M:%S'),
            *data
        ])

with open("test.csv", "w", encoding = 'UTF8', newline= '') as csv_file:
    writer = csv.writer(csv_file)
    writer.writerow(labels)
    writer.writerows(rows)

You don't need to use the csv library to write a csv file. Since the data is already comma delimited, you are just adding work.

labels = "date,time,s1,s2,s3,s4,s5\n"
samples = 20
rows = []

with Serial("COM3", 9600) as ser:
    for _ in range(samples):
        data = ser.readline()[:-2].decode('utf-8')
        now = datetime.now()
        rows.append("{},{},{}\n".format(
            now.strftime('%m/%d/%Y'),
            now.strftime('%H:%M:%S'),
            data))

with open("test.csv", "w", encoding = 'UTF8') as csv_file:
    csv_file.write(labels)
    csv_file.writelines(rows)

I don't think it makes sense to store all the data and then write it to the CSV file. Write the data as it is read. I would write your code like this:

labels = "date,time,s1,s2,s3,s4,s5\n"
samples = 20
with Serial("COM3", 9600) as ser:
    with open("test.csv", "w", encoding = 'UTF8', newline= '') as csv_file:
        csv_file.write(labels)
        for _ in range(samples):
            data = ser.readline()[:-2].decode('utf-8')
            now = datetime.now()
            row = "{},{},{}\n".format(
                now.strftime('%m/%d/%Y'),
                now.strftime('%H:%M:%S'),
                data)
            csv_file.write(row)

Why are you making separate columns for date and time. This makes date and time difficult to work with. Your date format is also a poor choice. I know month/day/year is a common convention, but it is worthless for sorting or comparison. I would have a single datetime column instead and use the default format.

labels = "time,s1,s2,s3,s4,s5\n"
samples = 20
with Serial("COM3", 9600) as ser:
    with open("test.csv", "w", encoding = 'UTF8', newline= '') as csv_file:
        csv_file.write(labels)
        for _ in range(samples):
            data = ser.readline()[:-2].decode('utf-8')
            csv_file.write(f"{datetime.now()},{data}\n")