infacted_prediction = []
hump_day = []
for name in names[1:]:
    max_num_infacteted = []
    hd = []
    for j in range(120):

        y = data[name]
        ind = np.where(y < 100)
        y = np.delete(y, ind)

        len_data = len(y)
        new_end = j+10
        if new_end > len_data-1:
            continue

        y = y[:new_end]

        len_data = len(y)
        x = np.arange(len_data)

        popt1, pcov = fit(func, x, y)

        x_off = popt1[1]

        x_fit = np.linspace(0, 120, 100)

        if False:
            pl.clf()
            pl.title(name)
            pl.plot(x,y, 'x')
            pl.plot(x_fit , func(x_fit, *popt1))
            pl.xlim([0, 50])
            pl.ylim([0, 200000])
            pl.pause(0.2)

        max_num_infacteted.append(func(100000, *popt1))
        hd.append(int(np.abs(popt1[1])))
    hump_day.append(hd)
    infacted_prediction.append(max_num_infacteted)


count = 0
for name in names[1:]:

    infected_pred = infacted_prediction[count][-1]
    infected_curr = data[name][-1]
    conver = np.nansum(np.gradient(infacted_prediction[count][-7:]))/7
    mean = np.nanmean(infacted_prediction[count][-3:])

    print(int(infected_curr),'\t', int(infected_pred),'\t', int(mean), '\t', int(np.abs(conver / len(infacted_prediction[count]))),'\t',name )
    count += 1

pl.clf()
count = 0
for name in names[1:]:

    pl.subplot(2,4,count+1)

    pl.title(name)

    pl.plot(infacted_prediction[count], label=name)

    if not ( count == 0 or count == 4 ):
        pl.yticks([])

    count += 1
    pl.xlim([0, 50])
    pl.ylim([0, 500000])

pl.show()

pl.clf()
count = 0
for name in names[1:]:

    pl.subplot(2,4,count+1)

    pl.title(name)

    #pl.plot(np.array(range(len(hump_day[count])))+10, hump_day[count], label=name)
    pl.plot(np.array(range(len(hump_day[count]))) + 10, np.abs(np.gradient(hump_day[count])), color='r', label=name)

    if not ( count == 0 or count == 4 ):
        pl.yticks([])

    count += 1
    pl.xlim([0, 50])
    pl.ylim([0, 20])

pl.show()