cauchyturing/imaging-time-series-to-improve-classification-and-imputation

Python 3.6+ (only)

imaging-time-series-to-improve-classification-and-imputation's Issues

AttributeError: 'Categorical' object has no attribute 'levels'

Getting error in line number 62 inside the serie21MTF.py file.
line 62 return np.array(MSM), label, q.levels
AttributeError: 'Categorical' object has no attribute 'levels'

need to use this code for generating MTF. Please help me

Compatibility issue

I have compatibility issue
Can you indicate which pandas and numpy versions you used?
Thx

what is the python version?

Hi, My python version is 3.4.4. when run your code ,there are error :
missing patentheses in call to print
print "in window_time_series",series

Question about the polor coordinate conversion

1. In the conversion to polor coordinate , the value is encoded as the angular cosine, and the angular is computed as in Eq (3) in your paper: \phi = arccos(xi), while in your released code, the angular is computed as phi = np.array(rescale(raw[k][1:])) * 2 * np.pi, it seems the arccos operation is missing?
2. in the function:

def paaMarkovMatrix(paalist, levels):
    # Generate Markov Matrix given a spesicif number of quantile bins
    paaindex = []
    for each in paalist:
        for level in levels:
            if level.left <= each <= level.right:
                paaindex.append(k)
    return paaindex

what does the variable 'k' means?

Guidance on Inverse Gramian Angular Field

Hi,

In the paper, your mention of the Gramian Angular Field and its unique inverse map in the polar coordinate system.

I was wondering if it would be possible for you to provide any code or guidance regarding the implementation of the inverse Gramian Angular Field.

Best regards,
Jiping

missing mpl_toolkits.axes_grid1

Hi, the code missing mpl_toolkits.axes_grid1

from mpl_toolkits.axes_grid1 import make_axes_locatable