[SciPy-user] timeseries and maskedarrays

[bevan]

Hello,

I recently posted on the numpy forum looking for a solution to creating totals 
and averages for time series of rainfall.  It was suggested to me that I try 
the timeseries package.

I have managed to get it working (struggling due to limited abilities rather 
than any issues with the package) - thanks by the way to the developers.

However,  when I run my code i get the following warning:
C:\Python25\lib\site-packages\scipy\sandbox\maskedarray\core.py:1521: 
UserWarning: Warning: converting a masked element to nan.
  warnings.warn("Warning: converting a masked element to nan.")

Is this an issue?
The main question I have is:
How would i create an average on values based on month, that is sum the 
rainfall from daily values to monthly totals (done), then average all the 
Januarys, Februarys etc in the timeseries?
Also any other issues with my code (below) that is likely to trip me up later?

Thanks for your time.


import time
import numpy
import maskedarray as MA
import timeseries as TS

def readTSFtimeseries(filename):
  """
  # PURPOSE: read a HYDROL .tsf file into a timeseries array.
  #
  # INPUT:
  #20/11/2003 @ 00:00:00                Q255 T5
  #21/11/2003 @ 00:00:00       0.000000 Q30  T5
  #............................................
  #07/05/2008 @ 00:00:00       2.500000 Q10  T5
  #
  # OUTPUT: a timeseries array
  #
  # EXAMPLE: >>> data = readTSFtimeseries(r'B:\Python\test.tsf')
  #          >>> x = data.dates        
  #          >>> y = data.series        
  """

  MYFILE = open(filename, "r")
  datafile = MYFILE.readlines()
  MYFILE.close()

  DateList=[]
  Rain_mmList=[]
  QualityList=[]
  TypeList=[]
  jd=[]

  for datalines in datafile:
      dataflds = datalines.split()
      datefld =time.strptime(dataflds[0]+'_'+dataflds[2], '%d/%m/%Y_%H:%M:%S')
      jd.append(datefld[7])
      DateList.append(TS.Date('D',year=datefld[0],month=datefld[1],day=datefld
[2]))

      if len(dataflds) == 6:
          Rain_mmList.append(float(dataflds[3]))
          QualityList.append(dataflds[4])
          TypeList.append(dataflds[5])

      if len(dataflds) == 5:
          #Rain_mmList.append(float(-9.99))
          Rain_mmList.append(TS.tsmasked)
          QualityList.append(dataflds[3])
          TypeList.append(dataflds[4])
          

  DateArr=TS.DateArray(dates=DateList,freq='D',copy =False)
  data=TS.time_series(Rain_mmList,DateArr)
  print data.dates.size,data.series.size
  return(data)


Raindata =readTSFtimeseries(r"C:\Documents and 
Settings\bevanj\Desktop\rain.tsf")

import pylab
import matplotlib
from timeseries import plotlib as TSPL

MonRaindata = Raindata.convert('monthly', MA.sum)
print MonRaindata
print Raindata.convert('yearly', MA.sum)
#fig1=TSPL.tsfigure()3
#fplt1=fig1.add_tsplot(111)
#fplt1.tsplot(RainData,'-')
#fplt1.tsplot(MonRainData,'-')
#pylab.show()