CODEWITHSUNDEEP
pythonpython-2.7pandas
weskpga
weskpga

Reputation: 2087

Ignore column length when adding columns to a DataFrame

I'm trying to add a bunch of returns to a DataFrame so that I can output it to a CSV file and do some basic graphing, but I'm getting an error message: ValueError: Length of values does not match length of index. I don't particularly care how long the columns are, but I was curious if it would be possible to assuage the index so that the first element in the list appears at index 0, and then everything else appears as subsequent values with NaN padding at the end if necessary. Is this doable in pandas?

for idx, col in enumerate(df.T.index.values):
    print "On " + str(idx) + " of " + str(len(df.T.index.values))
    startDate = dt.datetime.strptime(col.split()[0], "%m/%d/%y")
    endDate = startDate + dt.timedelta(days=91)

    returns1 = getReturnsForPeriod(df[col][1], startDate, endDate)
    retval[(col.split()[0], df[col][1], df[col][0])] = [ret for ret in returns1]

    returns2 = getReturnsForPeriod(df[col][2], startDate, endDate)
    retval[(col.split()[0], df[col][2], df[col][0])] = [ret for ret in returns2]

    returns3 = getReturnsForPeriod(df[col][3], startDate, endDate)
    retval[(col.split()[0], df[col][3], df[col][0])] = [ret for ret in returns3]

    returns4 = getReturnsForPeriod(df[col][4], startDate, endDate)
    retval[(col.split()[0], df[col][4], df[col][0])] = [ret for ret in returns4]

Upvotes: 1

Views: 679

Answers (1)

PlagTag
PlagTag

Reputation: 6439

I think this is a similar question to yours. Do it by going via dictionary first so you will know the final max length of your arrays.

Here just the basic mockup of the idea, not tested yet:

d = {}
for idx, col in enumerate(df.T.index.values):
  d.update(idx,col)
df = pandas.DataFrame.from_dict({idx,Series(col)) for idx,col in d.iteritems()})

Upvotes: 1

Related Questions