02:06:02	Christian Barreto:	# almacenamos en netcdfs
data_set = xr.Dataset( coords={"lon": (["lon"],sst_trop_jan_anom.lon),
                              "lat": (["lat"], sst_trop_jan_anom.lat)})

data_set["corr_pearson"] = (["lat", "lon"], m_corr_pearson)
data_set["corr_pearson_inf"] = (["lat", "lon"], m_corr_pearson_inf)
data_set["corr_pearson_sup"] = (["lat", "lon"], m_corr_pearson_sup)

data_set.to_netcdf("umbrales_de_corr.nc",mode="w")
02:24:48	Christian Barreto:	import numpy as np
import pandas as pd
import matplotlib.pyplot as plt

file="nino.xlsx"
nino=pd.read_excel(file,"NINO")
nino
02:30:05	Christian Barreto:	from statsmodels.graphics.tsaplots import plot_acf
fig1 = plt.figure()

plot_acf(nino12,lags=10)

plt.savefig("autocorr_nino12.png",dpi=300)
02:32:56	Christian Barreto:	fig1 = plt.figure()

plot_acf(nino34,lags=10)

plt.savefig("autocorr_nino34.png",dpi=300).
02:32:58	Christian Barreto:	fig1 = plt.figure()

plot_acf(nino34,lags=10)

plt.savefig("autocorr_nino34.png",dpi=300)
02:44:14	Christian Barreto:	########### con pandas
from pandas.plotting import autocorrelation_plot

autocorrelation_plot(nino34,label="nino34").set_xlim([0, 15])
autocorrelation_plot(nino12,label="nino12").set_xlim([0, 15])

plt.grid()
plt.legend()
02:50:36	Christian Barreto:	###################################
02:50:36	Christian Barreto:	import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import xarray as xr

dset = xr.open_dataset("sst.mnmean.nc")

sst=dset.sst

#seleccionamos tropicos

sst_trop=sst.sel(lat=slice(10,-10),lon=slice(0,360),time=slice("1981-01-01","2010-12-01"))
sst_trop.sel(time="2010-01-01").plot(robust=True,figsize=(10,3))
sst_trop
02:54:07	Christian Barreto:	############################
02:58:01	Christian Barreto:	m_autocorr_pearson = np.zeros((sst_trop.shape[1],sst_trop.shape[2]),dtype=float)

lagtime=3

def autocrr(x, t=lagtime):
    return np.corrcoef(np.array([x[:-t], x[t:]]))[0,1]
02:58:04	Christian Barreto:	m_autocorr_pearson = np.zeros((sst_trop.shape[1],sst_trop.shape[2]),dtype=float)

lagtime=3

def autocrr(x, t=lagtime):
    return np.corrcoef(np.array([x[:-t], x[t:]]))[0,1]

for ni in range(0,sst_trop.shape[2]): #long
    for nj in range(0,sst_trop.shape[1]): # loop lat
        
        sst_info=sst_trop.isel(lat=nj,lon=ni).values
        
        array_has_nan = np.isnan(np.sum(sst_info))
        
        if array_has_nan == True:
            
            m_autocorr_pearson[nj,ni]=np.nan
            
        else:
            m_autocorr_pearson[nj,ni] = autocrr(sst_info)
03:02:49	Christian Barreto:	https://wetransfer.com/downloads/b27b58b19cbe4817cf168f5cef25abf920210805233510/ef025d17513f774b8d3d46836b46c76320210805233529/5be287
03:08:16	Christian Barreto:	data_set2 = xr.Dataset( coords={'lon': ([ 'lon'], sst_trop.lon),
                                             'lat': (['lat',], sst_trop.lat)})

data_set2["corr_pearson"] = (["lat", "lon"], m_autocorr_pearson)

corr=data_set2.corr_pearson
corr.plot.contourf(robust=True,vmin=0.3,figsize=(10,3))
03:08:20	Lia Nicolle:	si ya descargó
03:08:57	ALANIA SUMARAN JOEL YOEL DZ02:	la mía falta aún :-(
03:10:56	Christian Barreto:	dset2 = xr.open_dataset("hgt.mon.mean.nc")

hgt=dset2.hgt

#seleccionamos tropicos

hgt_trop=hgt.sel(lat=slice(10,-10),lon=slice(0,360),time=slice("1981-01-01","2010-12-01"),level=200)
hgt_trop.sel(time="2010-01-01").plot(robust=True,figsize=(10,3))
hgt_trop
03:14:11	Christian Barreto:	m_autocorr_pearson = np.zeros((hgt_trop.shape[1],hgt_trop.shape[2]),dtype=float)

lagtime=3

def autocrr(x, t=lagtime):
    return np.corrcoef(np.array([x[:-t], x[t:]]))[0,1]

for ni in range(0,hgt_trop.shape[2]): #long
    for nj in range(0,hgt_trop.shape[1]): # loop lat
        
        hgt_info=hgt_trop.isel(lat=nj,lon=ni).values
        
        array_has_nan = np.isnan(np.sum(hgt_info))
        
        if array_has_nan == True:
            
            m_autocorr_pearson[nj,ni]=np.nan
            
        else:
            m_autocorr_pearson[nj,ni] = autocrr(hgt_info)
03:19:55	Christian Barreto:	data_set3 = xr.Dataset( coords={'lon': ([ 'lon'], hgt_trop.lon),
                                             'lat': (['lat',], hgt_trop.lat)})

data_set3["corr_pearson"] = (["lat", "lon"], m_autocorr_pearson)

corr2=data_set3.corr_pearson
corr2.plot.contourf(robust=True,vmin=0.3,figsize=(10,3))
03:23:54	Christian Barreto:	data_set3 = xr.Dataset( coords={'lon': ([ 'lon'], hgt_trop.lon),
                                             'lat': (['lat',], hgt_trop.lat)})

data_set3["corr_pearson"] = (["lat", "lon"], m_autocorr_pearson)

corr2=data_set3.corr_pearson
corr2.plot.contourf(robust=True,vmin=0.3,figsize=(10,3))
03:43:29	Christian Barreto:	https://we.tl/t-Kwoyu9bmid
03:47:33	Christian Barreto:	##############################
03:47:38	Christian Barreto:	import numpy as np
import pandas as pd
import matplotlib.pyplot as plt

file="datos_reg_lineal_esta.xlsx"
data=pd.read_excel(file)

data
03:49:09	Christian Barreto:	y=data["chuqui_jja"]
x=data["chuqui_mam"]
plt.scatter(x,y)
03:54:44	Lia Nicolle:	disculpe, podría aumentar un poco la letra por favor ?
03:55:49	Christian Barreto:	#create pandas para acumular datos

df= pd.DataFrame({"x":x,"y": y})

#calculamos la media de x e y

xmean= np.mean(x)
ymean= np.mean(y)

#calculamos los terminos que necesitamos en el numerador y denominador de beta
df["xycov"] = (df["x"] - xmean) * (df["y"] - ymean)
df["xvar"] = (df["x"] - xmean)**2

#calculamos a y b
b=df["xycov"].sum()/ df["xvar"].sum()
a= ymean-(b*xmean)
print("a:",a)
print("b:",b)
03:55:52	Christian Barreto:	##############
03:55:57	Christian Barreto:	ypred = a + b*x
plt.plot(y)
plt.plot(ypred)