import pandas as pd

import numpy as np

from scipy import interpolate

def interpDFrame( dfIn, tRFS1, tRFS2, min_time ):

# first column is time, others are dependencies on time

t_cycle = tRFS2 - tRFS1

t_marg = t_cycle * 0.0

max_time1 = min_time + t_cycle + t_marg

min_time1 = min_time - t_marg

# remove rows which are beyond full gait cycle

df1 = dfIn[(dfIn['time']>=min_time1)&(dfIn['time']<=max_time1)].copy()

time = df1['time'].values

t1 = (time-tRFS1)*100/t_cycle

df1['percentgaitcycle'] = pd.Series(t1, index=df1.index)

# modulo 100 percentgaitcycle

pm = np.mod(df1['percentgaitcycle'],100.0)

df1['pm'] = pd.Series(pm, index=df1.index)

# order by percentgaitcycle mod

df2 = df1.sort_values('pm')

pm = df2['pm'] # ordered pm

# we will interpolate non-time columns

df3 = df2.drop(['time','percentgaitcycle','pm'], axis=1)

cycleGrid = pd.Series(np.arange(0.0, 100.0))

cycleGrid.name = 'percentgaitcycle'

dfInterp=pd.DataFrame(cycleGrid)

for colName in df3.columns:

# interp function for x-y pair

interpFunc = interpolate.interp1d(pm, df3[colName], fill_value='extrapolate')

# calc values on the grid

colInterp = pd.Series(interpFunc(cycleGrid))

dfInterp[colName] = colInterp

return dfInterp