Issues Reading Dates from .txt File Python

Question

I currently have a script that I am using to plot part of a csv file after converting it to .txt. At this time, it works perfectly, except that when I change the dates in column 0 to ordinal form (I have done this so I can read all values as floats and perform calculations on column 4), Python chops off the hours, minutes and seconds. I still need the hours and minutes, because when I plot the data, it plots all of my points at the beginning of the day. Is there a way I can do this and keep the time as well as the date? I've tried converting the dates to a string and the other column to floats, but it's gotten very messy and confusing. Here is my code:

import numpy as np
import matplotlib.pyplot as plt
import matplotlib.dates as mdates
from datetime import datetime
import csv
from numpy import ma

def skip_first(seq,n):
    for i, item in enumerate(seq):
        if i >= n:
            yield item
g = open('soundTransit1_remote_rawMeasurements_15m.txt', 'w')
with open('soundTransit1_remote_rawMeasurements_15m.dat', 'rb') as f:
    csvreader = csv.reader(f)
    for row in skip_first(csvreader,4):
        for row in csv.reader(f,delimiter=',',skipinitialspace=True):
            print >>g, "\t".join(row)
g.close()

def date2str(date_str):
    date = datetime.strptime(date_str, '%Y-%m-%d %H:%M:%S')
    return date.toordinal()

def readfiles(file_list):
    """ read <TAB> delemited files as strings
        ignoring '# Comment' lines """
    data = []
    for fname in file_list:
        data.append(
                    np.loadtxt(fname,
                               usecols=(0,4),
                               comments='#',    # skip comment lines
                               delimiter='\t',
                               converters = { 0 : date2str },
                               dtype=None))
    return data

data = readfiles(['soundTransit1_remote_rawMeasurements_15m.txt'])
data_1 = ma.fix_invalid(data, fill_value = 'nan')

column_0 = np.array(data_1)[0][:,0]
airTempRaw = np.array(data_1)[0][:,1]

#Compute Air Temperature
airTempRs_ohms = 23100*(airTempRaw/(1-airTempRaw))
airTemp_degC = -39.17*np.log(airTempRs_ohms) + 410.43

def init_plot(title, yMin=-10, yMax=40):
    plt.figure(figsize=(24, 12))
    plt.title(title + disclamers)
    plt.xlabel(xtext)
    plt.ylabel(ytext)
    #plt.xlim(xMin,xMax)
    plt.ylim(yMin,yMax)
    plt.grid()
    #plt.xticks(np.arange(xMin,xMax+1))

def end_plot(name=None, cols=5):
    plt.legend(bbox_to_anchor=(0, -.1, 1, -0.5), loc=8, ncol=cols,
               mode="expand", borderaxespad=-1.,  scatterpoints=1)
    if name:
        plt.savefig(name, bbox_inches='tight')

disclamers = ('\nUSGS PROVISIONAL DATA'
          '\nSUBJECT TO REVISION'
          )
xtext = ('Date & Time')
ytext = ('Air Temperature, deg C')

init_plot('Air Temperature')

plt.plot(column_0, airTemp_degC, linestyle='-', color='b', label='Air Temperature')

plt.gca().xaxis.set_major_formatter(mdates.DateFormatter('%m/%d\n%H:%M'))
plt.gca().xaxis.set_minor_locator(mdates.HourLocator(interval=6))
plt.gca().xaxis.set_major_locator(mdates.DayLocator(interval=1))

end_plot(name='py_airTemp.png')

Thanks in advance for the help!

Have you considered storing the value in your CSV using a Unix Timestamp, and then just converting it to a readable time for your graph? This would give you a by-the-second resolution for your time storage, and still allow you to keep the date as well. — JPeroutek
– JPeroutek, Commented Jul 28, 2015 at 16:13
I haven't. I'm pretty new to Python, so I'm not sure how to go about doing that. — sfischer3394
– sfischer3394, Commented Jul 28, 2015 at 16:16
Check out this SO question. It has an example in the answers about getting the timestamp, and converting the timestamp to a readable date format. — JPeroutek
– JPeroutek, Commented Jul 28, 2015 at 16:17

Diziet Asahi · Accepted Answer · 2015-07-28 16:28:29Z

1

You don't say what is the format of the date column, but I think the problem lies with your converter in np.loadtxt().

Check this example from matplotlib: http://matplotlib.org/examples/pylab_examples/load_converter.html?highlight=strpdate2num

I believe this should work:

from matplotlib.dates import strpdate2num
def readfiles(file_list):
""" read <TAB> delemited files as strings
    ignoring '# Comment' lines """
data = []
for fname in file_list:
    data.append(
                np.loadtxt(fname,
                           usecols=(0,4),
                           comments='#',    # skip comment lines
                           delimiter='\t',
                           converters = { 0 : strpdate2num('%Y-%m-%d %H:%M:%S') },
                           dtype=None))
return data

answered Jul 28, 2015 at 16:28

Diziet Asahi

41k8 gold badges67 silver badges85 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

sfischer3394 Over a year ago

Beautiful!! That worked perfectly. Thank you so much!

Collectives™ on Stack Overflow

Issues Reading Dates from .txt File Python

1 Answer 1

1 Comment

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Linked

Related