#! /Usr/bin/python3

import scipy.optimize as opt
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import VisTools.plotting as vt
import uncertainties as unc
import uncertainties.unumpy as unp

import statsmodels.api as sm
from statsmodels.stats.outliers_influence import summary_table as st

plt.ioff()
plt.style.use('bmh')
fig, axarr = plt.subplots(
    2,
    1,
    figsize=(9, 8),
    gridspec_kw = {'height_ratios':[4,1]}
)
plt.subplots_adjust(hspace=.5)
data = pd.read_csv("./" + __file__[:-3].split("_")[1] + ".csv", comment="#")

fits = np.array([])
errs = np.array([])
colors = np.array([])

data['Dv'] = data['v_pit48/v'] - data['v_set/v']
data['Dvk'] = data['v_keith/v'] - data['v_set/v']
print(data)
# fit to abs dist
linfnc = lambda x,m,c: x*m+c

def plot_and_linfit(col: str,fil, l: str):
    p = axarr[0].errorbar(
        data[fil]['v_set/v'],
        data[fil][col],
        alpha = 0,
        label=None,
    #    fmt='.',
    #    antialiased=True
    )

    vals = vt.lm_plot(
        data[fil],
        'v_set/v',
        col,
        None,
        'dv_pit48/v',
        l,
        fig=axarr[0],
        color = p[0].get_color()
    )

    return vals, p

plot_and_linfit('Dvk', data.cycles == 0, 'ref')
for i in range(8):
    val, p = plot_and_linfit('Dv', data.cycles == i, 'scaler: {}'.format(i))
    fits = np.append(fits, val[0])
    #errs = np.append(errs, np.sqrt(pcov[0][0]))
    #colors = np.append(colors, p[0].get_color() )


print(fits[0].n)
for i in range(8):
    axarr[1].errorbar(
        i,
        fits[i].n,
        yerr = fits[i].s,
        fmt='.',
        #c = colors[i],
    )

axarr[0].set_xlabel('V$_{set}$/V')
#axarr[0].set_title("PowerIt ADC Calibration: dependency on measurement cycles")
axarr[0].set_ylabel('$\Delta$V$_{IN}$ / V')

axarr[1].set_xlabel('sampleTicks scaler')
axarr[1].set_ylabel('$\Delta(\Delta$V$_{IN})$')
axarr[0].legend(bbox_to_anchor=(0, -.28,1, 1), loc="lower left",
                mode="expand", borderaxespad=0, ncol=5)
#plt.tight_layout()
plt.savefig("./" + __file__[:-3].split("_")[1] + ".pdf", dpi=1000, bbox_inches='tight')


ss, d, ss2 = st(sm.OLS(data[data.cycles == 0]['v_pit48/v'], data[data.cycles == 0]['v_set/v']).fit(), alpha=.05)

low_ci, high_ci = d[:,4:6].T