0x0L

import pandas as pd

def pd_move_regress(X, Y, w, fit_intercept=True):
    assert (X.index == Y.index).all()
    B, A = move_regress(X.values, Y.values, w, fit_intercept=fit_intercept)
    
    idx = pd.MultiIndex.from_product([X.columns, Y.columns])
    B = pd.DataFrame(B.reshape(len(B), -1), index=X.index, columns=idx)
    A = pd.DataFrame(A, index=X.index, columns=Y.columns)
    return B, A

0x0L

from math import log, pi

import numba
import scipy.optimize as opt


@numba.jit(nopython=True, nogil=True)
def _precompute_dp(x, y):
    T = 1.0 * x.shape[0]
    

0x0L

# coding: utf-8

# In[ ]:


import numpy as np
import pandas as pd
import xarray as xr

0x0L

import io

def tf_summary_image(figure):
    figure.canvas.draw()
    w, h = figure.canvas.get_width_height()
    
    with io.BytesIO() as buffer:
        figure.savefig(buffer, format='png')
        image_str = buffer.getvalue()
    

0x0L

class MergeSequence(keras.utils.Sequence):
    def __init__(self, *seqs):
        lens = [len(s) for s in seqs]
        length = sum(lengths)
        
        indices = np.zeros((length, 2), dtype='int')
        for n, i in zip(lens, np.cumsum(lens))[:-1]:
            indices[i:] += [1, n]
            
        self.indices = indices

0x0L

from collections import defaultdict

import numpy as np
import pandas as pd
import tensorflow as tf
import xarray as xr

tf.enable_eager_execution()

0x0L

import numpy as np
import pandas as pd
import xarray as xr


@xr.register_dataarray_accessor('str')
class StringAccessor:
    def __init__(self, xarray_obj):
        self._obj = xarray_obj
        self._data = xarray_obj.data

0x0L

public static class Pandas
{
    public static double[] Ewma(
        double[] vals, double com,
        bool adjust=true, bool ignore_na=false, int minp=0)
    {
        int N = vals.Length;
        var output = new double[N];

        if (N == 0)

0x0L

def diag(A, k=0):
    s = slice(
        max(k, -A.shape[1] * k),
        max(0, (A.shape[1] - k) * A.shape[1]),
        A.shape[1] + 1
    )
    return A.ravel()[s]

0x0L