Analysing data from weights and biases

Import relevant tools

[2]:

# Convenient jupyter setup
%load_ext autoreload
%autoreload 2
%config IPCompleter.greedy=True

The autoreload extension is already loaded. To reload it, use:
  %reload_ext autoreload

[3]:

# set up plotting settings.
from src.plot_utils import ps_defaults
ps_defaults()

[3]:

import numpy as np
from src.wandb_utils import metric_conv_data
from src.models.poly import fit
import matplotlib.pyplot as plt
from uncertainties import unumpy as unp

 Collecting data from wandb

Collect all the required data from a particular run with other paramters controlled.

[4]:

mem_dict = {}
for mem in ["EEEE", "EECE", "EEEC", "EECC"]:
    metric_d, _ = metric_conv_data(
        metric_name="trend_nino3.4",
        prefix="k_days_10_eps_days",
        ex_list=["ingrid_True"],
        control_variable_list=[
            (("atm", "k_days"), 10),
            # (("coup", "c_d"), 2.25e-3),
            (("atm", "e_frac"), 0.5),
            (("atm", "eps_days"), 0.75),
            (("atm", "mem"), mem),
            (("atm", "vary_cloud_const"), True),
        ],
        index_by=("coup", "c_d"),
    )
    mem_dict[mem] = [[], []]
    for val in metric_d:
        mem_dict[mem][0].append(float(val))
        mem_dict[mem][1].append(float(metric_d[val][5, 1]))

k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00025_vary_cloud_const_True_EEEE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0_vary_cloud_const_True_EEEE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.002_vary_cloud_const_True_EEEE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.001_vary_cloud_const_True_EEEE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.003_vary_cloud_const_True_EEEE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.002_vary_cloud_const_True_EEEE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00125_vary_cloud_const_True_EEEE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00175_vary_cloud_const_True_EEEE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.0015_vary_cloud_const_True_EEEE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_True_EEEE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.0025_vary_cloud_const_True_EEEE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00025_vary_cloud_const_True_EECE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0_vary_cloud_const_True_EECE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00075_vary_cloud_const_True_EECE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.001_vary_cloud_const_True_EECE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.003_vary_cloud_const_True_EECE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.002_vary_cloud_const_True_EECE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00125_vary_cloud_const_True_EECE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00275_vary_cloud_const_True_EECE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00175_vary_cloud_const_True_EECE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.0015_vary_cloud_const_True_EECE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_True_EECE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.0025_vary_cloud_const_True_EECE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.0005_vary_cloud_const_True_EEEC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00025_vary_cloud_const_True_EEEC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0_vary_cloud_const_True_EEEC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.001_vary_cloud_const_True_EEEC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.003_vary_cloud_const_True_EEEC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.002_vary_cloud_const_True_EEEC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00125_vary_cloud_const_True_EEEC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00175_vary_cloud_const_True_EEEC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.0015_vary_cloud_const_True_EEEC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_True_EEEC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.0025_vary_cloud_const_True_EEEC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.0005_vary_cloud_const_True_EECC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00025_vary_cloud_const_True_EECC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.002_vary_cloud_const_True_EECC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0_vary_cloud_const_True_EECC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.001_vary_cloud_const_True_EECC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00075_vary_cloud_const_True_EECC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.003_vary_cloud_const_True_EECC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.002_vary_cloud_const_True_EECC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00125_vary_cloud_const_True_EECC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00275_vary_cloud_const_True_EECC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00175_vary_cloud_const_True_EECC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.0015_vary_cloud_const_True_EECC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_True_EECC
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.0025_vary_cloud_const_True_EECC

Plot data from wandb

[5]:

color_d = {
    "EEEE": "blue",
    "EECE": "green",
    "EEEC": "orange",
    "EECC": "red",
}

name_d = {
    "EEEE": "ECMWF",
    "EECE": "W",
    "EEEC": "RH",
    "EECC": "RH+W",
}
min_x = np.inf
max_x = -np.inf

for mem in name_d:
    if min(mem_dict[mem][0]) < min_x:
        min_x = min(mem_dict[mem][0])
    if max(mem_dict[mem][0]) > max_x:
        max_x = max(mem_dict[mem][0])

ext = 0.05
min_x_pred = min_x - (max_x - min_x) * ext
max_x_pred = max_x + (max_x - min_x) * ext
x_pred = np.linspace(min_x_pred, max_x_pred, num=50)

for mem in name_d:
    param, func = fit(mem_dict[mem][0], mem_dict[mem][1], reg_type="lin")
    print("param", mem, param)
    y_pred = func(x_pred)
    y_pred_n = unp.nominal_values(y_pred)
    y_pred_s = unp.std_devs(y_pred)
    plt.fill_between(
        x_pred,
        y_pred_n + y_pred_s,
        y_pred_n - y_pred_s,
        alpha=0.25,
        color=color_d[mem],
    )
    plt.plot(x_pred, y_pred_n, color=color_d[mem], alpha=0.5)
    plt.scatter(
        mem_dict[mem][0],
        mem_dict[mem][1],
        marker="x",
        label=name_d[mem],
        c=color_d[mem],
    )

plt.xlabel("Drag coefficient, $C_d$, [dimesionless]")
plt.ylabel("1958-2017 nino3.4 trend [K]")
plt.legend(
    bbox_to_anchor=(0.0, 1.02, 1, 0.102),
    loc="lower left",
    mode="expand",
    ncol=4,
)
plt.gca().ticklabel_format(axis="x", style="sci", scilimits=(0, 0), useMathText=True)

plt.xlim(min_x_pred, max_x_pred)
plt.tight_layout()

param EEEE [-20.01127415639631+/-0.605798369381671
 0.4810086265055368+/-0.001087909985178536]
param EECE [-19.664065317776124+/-0.6014229809489599
 0.5366611332943885+/-0.0010387315828538726]
param EEEC [-6.258569960543675+/-0.9998655685643177
 0.7959540076231921+/-0.001718648144296277]
param EECC [6.530101974397601+/-0.7340304753418364
 0.9315313394718199+/-0.001297594806424033]
../_images/notebooks_analyse_results_9_1.png 
[6]:

_[0].cfg["oc"].nummode

[6]:

2

[7]:

_[0.00225].cfg["oc"].nummode

[7]:

2

Arrow plots

[8]:

mem_dict = {}
for mem in ["EEEE", "EECE", "EEEC", "EECC"]:
    metric_d = {}
    for vary_cloud_const in [True, False]:
        for e_frac in [0.5, 2]:
            metric_d_loc, _ = metric_conv_data(
                metric_name="trend_nino3.4",
                prefix="",
                ex_list=["ingrid_True"],
                control_variable_list=[
                    (("atm", "k_days"), 10),
                    (("coup", "c_d"), 2.25e-3),
                    (("atm", "e_frac"), e_frac),
                    (("atm", "eps_days"), 0.75),
                    (("atm", "mem"), mem),
                    (("atm", "vary_cloud_const"), vary_cloud_const),
                ],
                index_by=("name"),
            )
            metric_d = {**metric_d_loc, **metric_d}

    mem_dict[mem] = [[], []]
    for val in metric_d:
        mem_dict[mem][0].append(val)
        mem_dict[mem][1].append(float(metric_d[val][5, 1]))

test25
test24
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_True_EEEE
N_ECMWF_coup
ECMWF_coup
E2_ECMWF_coup
E_ECMWF_coup
A_ECMWF_coup
AE_ECMWF_coup
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_True_EECE
N_C_W2_coup
N_C_W_coup
C_W_coup
E2_C_W_coup
E_C_W_coup
A_C_W2_coup
A_C_W_coup
AE_C_W_coup
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_True_EEEC
N_C_RH_coup
C_RH_coup
E2_C_RH_coup
E_C_RH_coup
A_C_RH_coup
AE_C_RH_coup
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_True_EECC
N_C_RH_W2_coup
N_C_RH_W_coup
C_RH_W_coup
E2_C_RH_W_coup
E_C_RH_W_coup
A_C_RH_W2_coup
A_C_RH_W_coup
AE_C_RH_W_coup

[9]:

i = 1
for mem in mem_dict:
    #print([i for _ in mem_dict[mem][1]], mem_dict[mem][1])
    plt.scatter([i for _ in mem_dict[mem][1]], mem_dict[mem][1], label=mem)
    i += 1

plt.xlim(0, i)
plt.legend()

[9]:

<matplotlib.legend.Legend at 0x7f8beafccd60>
../_images/notebooks_analyse_results_14_1.png 
[10]:

metric_d

[10]:

{'AE_C_RH_W_coup': array([[0.        , 0.55993331],
        [1.        , 0.46083996],
        [2.        , 0.74789722],
        [3.        , 0.77034592],
        [4.        , 0.82005772],
        [5.        , 0.82740342]]),
 'A_C_RH_W2_coup': array([[0.        , 0.55993331],
        [1.        , 0.40156689],
        [2.        , 0.71137339],
        [3.        , 0.72161181],
        [4.        , 0.77504443],
        [5.        , 0.78024562]]),
 'A_C_RH_W_coup': array([[0.        , 0.55993331],
        [1.        , 0.35873489],
        [2.        , 0.63544161],
        [3.        , 0.60192197],
        [4.        , 0.65120501],
        [5.        , 0.6474647 ]]),
 'E2_C_RH_W_coup': array([[0.        , 0.55993331],
        [1.        , 0.52378907],
        [2.        , 0.81098679],
        [3.        , 0.92877046],
        [4.        , 0.97999571],
        [5.        , 1.00194651]]),
 'E_C_RH_W_coup': array([[0.        , 0.55993331],
        [1.        , 0.35293816],
        [2.        , 0.59224832],
        [3.        , 0.57863006],
        [4.        , 0.61877396],
        [5.        , 0.61799476]]),
 'k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_True_EECC': array([[0.        , 0.55993331],
        [1.        , 0.46157158],
        [2.        , 0.77268467],
        [3.        , 0.87324829],
        [4.        , 0.92897589],
        [5.        , 0.94773791]]),
 'N_C_RH_W2_coup': array([[0.        , 0.55993331],
        [1.        , 0.46157158],
        [2.        , 0.77268467],
        [3.        , 0.87324829],
        [4.        , 0.92897589],
        [5.        , 0.94773791]]),
 'N_C_RH_W_coup': array([[0.        , 0.55993331],
        [1.        , 0.4071576 ],
        [2.        , 0.68321311],
        [3.        , 0.72376622],
        [4.        , 0.77531703],
        [5.        , 0.78381075]]),
 'C_RH_W_coup': array([[0.        , 0.55993331],
        [1.        , 0.3019285 ],
        [2.        , 0.56477079],
        [3.        , 0.5379757 ],
        [4.        , 0.58224876],
        [5.        , 0.57972549]])}

[35]:

from src.search import variable_combinations
variable_combinations(vary=[True, True, True, True], exps=["6"])
mem_dict = {}
for mem in variable_combinations(vary=[True, True, True, True], exps=["6"]):
    #["EEEE", "6EEE", "E6EE", "EE6E", "E66E", "EEE6", "EE66", "E666", "6666"]:
    metric_d = {}
    for vary_cloud_const in [True, False]:
        for e_frac in [0.5, 2]:
            metric_d_loc, _ = metric_conv_data(
                metric_name="trend_nino3.4",
                prefix="",
                ex_list=["ingrid_True"],
                control_variable_list=[
                    (("atm", "k_days"), 10),
                    (("coup", "c_d"), 2.25e-3),
                    (("atm", "e_frac"), e_frac),
                    (("atm", "eps_days"), 0.75),
                    (("atm", "mem"), mem),
                    (("atm", "vary_cloud_const"), vary_cloud_const),
                ],
                index_by=("name"),
            )
            metric_d = {**metric_d_loc, **metric_d}

    mem_dict[mem] = [[], []]
    for val in metric_d:
        mem_dict[mem][0].append(val)
        mem_dict[mem][1].append(float(metric_d[val][5, 1]))

i = 1
for mem in mem_dict:
    #print([i for _ in mem_dict[mem][1]], mem_dict[mem][1])
    plt.scatter([i for _ in mem_dict[mem][1]], mem_dict[mem][1], label=mem)
    i += 1

plt.xlim(0, i)
plt.legend()

test25
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_True_EEEE
N_ECMWF_coup
ECMWF_coup
E2_ECMWF_coup
E_ECMWF_coup
A_ECMWF_coup
AE_ECMWF_coup
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_True_EEE6
k_days_10_eps_days_0.75_e_frac_2_c_d_0.00225_vary_cloud_const_True_EEE6
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_False_EEE6
k_days_10_eps_days_0.75_e_frac_2_c_d_0.00225_vary_cloud_const_False_EEE6
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_True_6E6E
k_days_10_eps_days_0.75_e_frac_2_c_d_0.00225_vary_cloud_const_True_6E6E
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_False_6E6E
k_days_10_eps_days_0.75_e_frac_2_c_d_0.00225_vary_cloud_const_False_6E6E
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_True_E66E
k_days_10_eps_days_0.75_e_frac_2_c_d_0.00225_vary_cloud_const_True_E66E
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_False_E66E
k_days_10_eps_days_0.75_e_frac_2_c_d_0.00225_vary_cloud_const_False_E66E
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_True_6EEE
k_days_10_eps_days_0.75_e_frac_2_c_d_0.00225_vary_cloud_const_True_6EEE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_False_6EEE
k_days_10_eps_days_0.75_e_frac_2_c_d_0.00225_vary_cloud_const_False_6EEE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_True_EE6E
k_days_10_eps_days_0.75_e_frac_2_c_d_0.00225_vary_cloud_const_True_EE6E
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_False_EE6E
k_days_10_eps_days_0.75_e_frac_2_c_d_0.00225_vary_cloud_const_False_EE6E
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_True_E6EE
k_days_10_eps_days_0.75_e_frac_2_c_d_0.00225_vary_cloud_const_True_E6EE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_False_E6EE
k_days_10_eps_days_0.75_e_frac_2_c_d_0.00225_vary_cloud_const_False_E6EE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_True_E6E6
k_days_10_eps_days_0.75_e_frac_2_c_d_0.00225_vary_cloud_const_True_E6E6
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_False_E6E6
k_days_10_eps_days_0.75_e_frac_2_c_d_0.00225_vary_cloud_const_False_E6E6
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_True_66EE
k_days_10_eps_days_0.75_e_frac_2_c_d_0.00225_vary_cloud_const_True_66EE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_False_66EE
k_days_10_eps_days_0.75_e_frac_2_c_d_0.00225_vary_cloud_const_False_66EE
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_True_EE66
k_days_10_eps_days_0.75_e_frac_2_c_d_0.00225_vary_cloud_const_True_EE66
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_False_EE66
k_days_10_eps_days_0.75_e_frac_2_c_d_0.00225_vary_cloud_const_False_EE66
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_True_E666
k_days_10_eps_days_0.75_e_frac_2_c_d_0.00225_vary_cloud_const_True_E666
k_days_10_eps_days_0.75_e_frac_0.5_c_d_0.00225_vary_cloud_const_False_E666
k_days_10_eps_days_0.75_e_frac_2_c_d_0.00225_vary_cloud_const_False_E666

[35]:

<matplotlib.legend.Legend at 0x7f8b631c69d0>
../_images/notebooks_analyse_results_16_2.png 
[51]:

i = 1
xticks = []
xticklabels = []
mem_keys = sorted(mem_dict.keys())
for mem in mem_keys:
    # print([i for _ in mem_dict[mem][1]], mem_dict[mem][1])
    if len(mem_dict[mem][1]) != 0:
        plt.scatter(i, np.mean(mem_dict[mem][1]), color="black")
        plt.errorbar(
            i,
            np.mean(mem_dict[mem][1]),
            capsize=3,
            ecolor="grey",
            fmt="+",
            color="black",
            yerr=np.std(mem_dict[mem][1]),
        )
        plt.scatter([i for _ in mem_dict[mem][1]], mem_dict[mem][1], label=mem)
        xticks.append(i)
        xticklabels.append(mem + "\n (" + str(len(mem_dict[mem][1])) + ")")
        i += 1

plt.xticks(xticks, xticklabels)
plt.xlim(0, i)
# plt.legend()

[51]:

(0.0, 12.0)
../_images/notebooks_analyse_results_17_1.png 
[40]:

mem_dict.keys()

[40]:

dict_keys(['EEEE', 'EEE6', '6E6E', '6EE6', 'E66E', '6EEE', 'EE6E', '6E66', 'E6EE', '666E', 'E6E6', '66EE', '66E6', '6666', 'EE66', 'E666'])

[48]:

mem_keys = sorted(mem_dict.keys())
mem_keys.reverse()

[88]:

from src.models.model_setup import mems_to_df
df = mems_to_df(mem_keys)

[100]:

fig, ax = plt.subplots()
pd.plotting.table(ax,df.transpose())
plt.xticks([])
plt.yticks([])
# Hide grid lines
ax.grid(False)
plt.axis('off')
plt.tight_layout()
../_images/notebooks_analyse_results_21_0.png 
[52]:

mem_keys

[52]:

['6666',
 '666E',
 '66E6',
 '66EE',
 '6E66',
 '6E6E',
 '6EE6',
 '6EEE',
 'E666',
 'E66E',
 'E6E6',
 'E6EE',
 'EE66',
 'EE6E',
 'EEE6',
 'EEEE']

[60]:

from src.constants import MODEL_NAMES, VAR_DICT
def mem_to_summary(mem:str) -> dict:
    result_d = {}
    for i, value in enumerate(mem):
        print(i, VAR_DICT[i], MODEL_NAMES[value])
        result_d[VAR_DICT[i]] = MODEL_NAMES[value]
    return result_d
mem_to_summary("EEEE")
mem_to_summary("EEE6")
mem_to_summary("6EEC")

0 ts ECMWF
1 clt ECMWF
2 sfcWind ECMWF
3 rh ECMWF
0 ts ECMWF
1 clt ECMWF
2 sfcWind ECMWF
3 rh CMIP6
0 ts CMIP6
1 clt ECMWF
2 sfcWind ECMWF
3 rh CMIP5

[60]:

{'ts': 'CMIP6', 'clt': 'ECMWF', 'sfcWind': 'ECMWF', 'rh': 'CMIP5'}

Summary tables of runs

[3]:

from src.wandb_utils import summary_table

[4]:

summary_table("sdat2/ENSOTrend-gamma")

[4]:
index ts clt sfcWind rh pr ps tau c_d eps_days eps_frac vary_cloud_const trend_nino3.4 [degC] mean_nino3.4 [degC] mean_pac [degC]
0 SESS CMIP6 ECMWF CMIP6 CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 2.0 False 0.505120 27.087601 26.218040
1 EECCE ECMWF ECMWF CMIP5 CMIP5 ECMWF ECMWF ECMWF 0.00225 0.75 2.0 False 0.817237 27.257566 26.300879
2 ESSE ECMWF CMIP6 CMIP6 ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 2.0 False 0.335936 27.076225 26.212515
3 SESS CMIP6 ECMWF CMIP6 CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 2.0 True 0.611463 27.136320 26.233788
4 EECCE ECMWF ECMWF CMIP5 CMIP5 ECMWF ECMWF ECMWF 0.00225 0.75 2.0 True 1.064891 27.276243 26.313169
5 ESSE ECMWF CMIP6 CMIP6 ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 2.0 True 0.000000 0.000000 1.024149
6 SESS CMIP6 ECMWF CMIP6 CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 1.0 False 0.485317 27.078705 26.216551
7 EECCE ECMWF ECMWF CMIP5 CMIP5 ECMWF ECMWF ECMWF 0.00225 0.75 1.0 False 0.785651 27.256149 26.300472
8 ESSE ECMWF CMIP6 CMIP6 ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 1.0 False 0.318892 27.068947 26.211222
9 SESS CMIP6 ECMWF CMIP6 CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 1.0 True 0.588602 27.125921 26.231976
10 EECCE ECMWF ECMWF CMIP5 CMIP5 ECMWF ECMWF ECMWF 0.00225 0.75 1.0 True 1.025281 27.274645 26.312494
11 ESSE ECMWF CMIP6 CMIP6 ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 1.0 True 0.000000 0.000000 1.024149
12 EECCE ECMWF ECMWF CMIP5 CMIP5 ECMWF ECMWF ECMWF 0.00225 0.75 0.5 False 0.765463 27.254919 26.300161
13 ESSE ECMWF CMIP6 CMIP6 ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 0.5 False 0.308752 27.064651 26.210520
14 SESS CMIP6 ECMWF CMIP6 CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 0.5 False 0.473950 27.073715 26.215836
15 EEEEE ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 0.5 True 0.408055 27.134361 26.235750
16 EECCE ECMWF ECMWF CMIP5 CMIP5 ECMWF ECMWF ECMWF 0.00225 0.75 0.5 True 0.999500 27.273149 26.312008 
[5]:

summary_table("sdat2/ENSOTrend-beta")

[5]:
index ts clt sfcWind rh pr ps tau c_d eps_days eps_frac vary_cloud_const trend_nino3.4 [degC] mean_nino3.4 [degC] mean_pac [degC]
0 EEGG ECMWF ECMWF MOHC.HadGEM3-GC31-LL.r1i1p1f3 MOHC.HadGEM3-GC31-LL.r1i1p1f3 ECMWF ECMWF ECMWF 0.00225 0.75 0.5 True 0.000000 0.000000 5.633691
1 SESS CMIP6 ECMWF CMIP6 CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 0.5 True 0.575857 27.120268 26.231049
2 SEES CMIP6 ECMWF ECMWF CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 0.5 True 0.573042 27.119400 26.248104
3 SESE CMIP6 ECMWF CMIP6 ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 0.5 True 0.426699 27.060863 26.211390
4 ESSE ECMWF CMIP6 CMIP6 ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 0.5 True 0.000000 0.000000 1.024149
5 ESSS ECMWF CMIP6 CMIP6 CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 0.5 True 0.506604 27.138105 26.234684
6 ESES ECMWF CMIP6 ECMWF CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 0.5 True 0.509718 27.141045 26.252346
7 EECE ECMWF ECMWF CMIP5 ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 2.0 False 0.408031 27.107105 26.239676
8 EEEC ECMWF ECMWF ECMWF CMIP5 ECMWF ECMWF ECMWF 0.00225 0.75 2.0 True 0.807164 27.264380 26.302011
9 EECC ECMWF ECMWF CMIP5 CMIP5 ECMWF ECMWF ECMWF 0.00225 0.75 2.0 False 0.771743 27.224466 26.298929
10 EECE ECMWF ECMWF CMIP5 ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 2.0 True 0.494428 27.145832 26.251301
11 EEEC ECMWF ECMWF ECMWF CMIP5 ECMWF ECMWF ECMWF 0.00225 0.75 2.0 False 0.625908 27.182840 26.283918
12 EECC ECMWF ECMWF CMIP5 CMIP5 ECMWF ECMWF ECMWF 0.00225 0.75 2.0 True 0.994551 27.322180 26.323425
13 EECE ECMWF ECMWF CMIP5 ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 1.0 False 0.386119 27.097765 26.238108
14 EESE ECMWF ECMWF CMIP6 ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 2.0 False 0.355958 27.086599 26.210655
15 EEEC ECMWF ECMWF ECMWF CMIP5 ECMWF ECMWF ECMWF 0.00225 0.75 1.0 True 0.771080 27.248077 26.299635
16 EECC ECMWF ECMWF CMIP5 CMIP5 ECMWF ECMWF ECMWF 0.00225 0.75 1.0 False 0.735869 27.208372 26.296741
17 EECE ECMWF ECMWF CMIP5 ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 1.0 True 0.470243 27.135424 26.249355
18 SSSS CMIP6 CMIP6 CMIP6 CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 2.0 True 0.576529 27.119604 26.233866
19 EESE ECMWF ECMWF CMIP6 ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 2.0 True 0.427607 27.118626 26.219982
20 EEEC ECMWF ECMWF ECMWF CMIP5 ECMWF ECMWF ECMWF 0.00225 0.75 0.5 True 0.594246 27.168831 26.282095
21 EECC ECMWF ECMWF CMIP5 CMIP5 ECMWF ECMWF ECMWF 0.00225 0.75 1.0 True 0.952061 27.302366 26.320534
22 SSSS CMIP6 CMIP6 CMIP6 CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 2.0 False 0.463512 27.067633 26.217897
23 EEES ECMWF ECMWF ECMWF CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 2.0 False 0.477161 27.126596 26.243036
24 EEES ECMWF ECMWF ECMWF CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 2.0 True 0.586770 27.175640 26.255524
25 EESE ECMWF ECMWF CMIP6 ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 1.0 False 0.336697 27.078407 26.209253
26 EECE ECMWF ECMWF CMIP5 ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 0.5 True 0.371520 27.091593 26.237110
27 EEEE ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 2.0 True 0.435412 27.122578 26.236391
28 EEEC ECMWF ECMWF ECMWF CMIP5 ECMWF ECMWF ECMWF 0.00225 0.75 0.5 False 0.574338 27.160229 26.281122
29 SSSS CMIP6 CMIP6 CMIP6 CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 1.0 True 0.554943 27.109760 26.232132
30 EEES ECMWF ECMWF ECMWF CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 1.0 False 0.450163 27.114744 26.241316
31 EEES ECMWF ECMWF ECMWF CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 1.0 True 0.556484 27.162172 26.253450
32 EESE ECMWF ECMWF CMIP6 ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 1.0 True 0.406319 27.109489 26.218246
33 EECC ECMWF ECMWF CMIP5 CMIP5 ECMWF ECMWF ECMWF 0.00225 0.75 0.5 True 0.711993 27.197847 26.295433
34 EEEE ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 2.0 False 0.360966 27.089426 26.228193
35 SSSS CMIP6 CMIP6 CMIP6 CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 1.0 False 0.445372 27.059433 26.216509
36 EEES ECMWF ECMWF ECMWF CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 0.5 False 0.432872 27.107285 26.240362
37 EEES ECMWF ECMWF ECMWF CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 0.5 True 0.537050 27.153660 26.252243
38 EESE ECMWF ECMWF CMIP6 ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 0.5 False 0.324840 27.073420 26.208481
39 EESS ECMWF ECMWF CMIP6 CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 2.0 True 0.588144 27.174639 26.238489
40 EEEE ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 1.0 True 0.414116 27.113482 26.234781
41 SSSS CMIP6 CMIP6 CMIP6 CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 0.5 True 0.543124 27.104477 26.231276
42 EESE ECMWF ECMWF CMIP6 ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 0.5 True 0.393731 27.104143 26.217299
43 EESS ECMWF ECMWF CMIP6 CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 1.0 True 0.557512 27.160902 26.236244
44 EEEE ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 0.5 True 0.341209 27.081051 26.226826
45 SSSS CMIP6 CMIP6 CMIP6 CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 0.5 False 0.435101 27.054873 26.215855
46 EESS ECMWF ECMWF CMIP6 CMIP6 ECMWF ECMWF ECMWF 0.00225 0.75 0.5 True 0.539075 27.152735 26.235046
47 EEEE ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 0.5 False 0.328922 27.075922 26.226074
48 EEEEEEE ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 0.5 True 0.400531 27.107738 26.233780 
[4]:

summary_table("sdat2/seager19")

Missing key mem
    full_key: atm.mem
    object_type=dict
Missing key mem
    full_key: atm.mem
    object_type=dict
Missing key mem
    full_key: atm.mem
    object_type=dict
Missing key mem
    full_key: atm.mem
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key vary_cloud_const
    full_key: atm.vary_cloud_const
    object_type=dict
Missing key c_d
    full_key: coup.c_d
    object_type=dict
Missing key c_d
    full_key: coup.c_d
    object_type=dict
Missing key c_d
    full_key: coup.c_d
    object_type=dict
Missing key c_d
    full_key: coup.c_d
    object_type=dict
Missing key c_d
    full_key: coup.c_d
    object_type=dict
Missing key c_d
    full_key: coup.c_d
    object_type=dict
Missing key c_d
    full_key: coup.c_d
    object_type=dict
Missing key c_d
    full_key: coup.c_d
    object_type=dict
Missing key c_d
    full_key: coup.c_d
    object_type=dict
Missing key c_d
    full_key: coup.c_d
    object_type=dict

[4]:
index ts clt sfcWind rh pr ps tau c_d eps_days eps_frac vary_cloud_const trend_nino3.4 [degC] mean_nino3.4 [degC] mean_pac [degC]
0 6EEE CMIP6 ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 2.0 False 0.406436 27.104700 2.622433e+01
1 66EE CMIP6 CMIP6 ECMWF ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 2.0 False 0.376009 27.089785 2.622624e+01
2 6E6E CMIP6 ECMWF CMIP6 ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 2.0 False 0.474587 27.132683 2.629025e+01
3 66EE CMIP6 CMIP6 ECMWF ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 2.0 True -26905.620581 -5602.992188 -1.569812e+14
4 6EEE CMIP6 ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 2.0 True 0.467564 27.132328 2.623201e+01
... ... ... ... ... ... ... ... ... ... ... ... ... ... ... ...
154 EEEE ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 2.0 True 0.366095 27.084520 2.619155e+01
155 EEEE ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 0.5 True 0.343836 27.074650 2.618952e+01
156 EEEf ECMWF ECMWF ECMWF fixed ECMWF ECMWF ECMWF 0.00225 0.75 0.5 False 0.397791 27.106426 2.625883e+01
157 EEEE ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 2.0 False 0.400535 27.101734 2.622359e+01
158 EEEE ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF 0.00225 0.75 0.5 False 0.377373 27.091391 2.622142e+01

159 rows × 15 columns

[9]:

from src.wandb_utils import aggregate_table

[11]:

cmip5_table = aggregate_table(project="sdat2/ENSOTrend-gamma", mem_list=["EEEE", "EECE", "EEEC", "EECC"])
cmip5_table

'summary_table'  1.62282 s

'_aggregate_matches'  0.00867 s


[11]:
ts clt sfcWind rh pr ps tau trend_nino3.4 [degC] mean_nino3.4 [degC] mean_pac [degC] N
EEEE ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF 0.41+/-0.04 27.131+/-0.007 26.2345+/-0.0022 6
EECE ECMWF ECMWF CMIP5 ECMWF ECMWF ECMWF ECMWF 0.46+/-0.04 27.151+/-0.009 26.247+/-0.004 6
EEEC ECMWF ECMWF ECMWF CMIP5 ECMWF ECMWF ECMWF 0.73+/-0.10 27.238+/-0.011 26.292+/-0.004 6
EECC ECMWF ECMWF CMIP5 CMIP5 ECMWF ECMWF ECMWF 0.91+/-0.13 27.265+/-0.010 26.307+/-0.007 6 
[12]:

cmip6_table = aggregate_table(
    project="sdat2/ENSOTrend-gamma",
    mem_list=["EEEE", "EESE", "EEES", "EESS", "SSSSSSE"],
)
cmip6_table

'summary_table'  1.73339 s

'_aggregate_matches'  0.01099 s


[12]:
ts clt sfcWind rh pr ps tau trend_nino3.4 [degC] mean_nino3.4 [degC] mean_pac [degC] N
EEEE ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF ECMWF 0.41+/-0.04 27.131+/-0.007 26.2345+/-0.0022 6
EESE ECMWF ECMWF CMIP6 ECMWF ECMWF ECMWF ECMWF 0.40+/-0.04 27.127+/-0.007 26.2173+/-0.0029 6
EEES ECMWF ECMWF ECMWF CMIP6 ECMWF ECMWF ECMWF 0.54+/-0.06 27.174+/-0.010 26.250+/-0.004 6
EESS ECMWF ECMWF CMIP6 CMIP6 ECMWF ECMWF ECMWF 0.54+/-0.06 27.169+/-0.009 26.232+/-0.004 6
SSSSSSE CMIP6 CMIP6 CMIP6 CMIP6 CMIP6 CMIP6 ECMWF 0.58+/-0.06 27.108+/-0.014 26.225+/-0.006 6

Arrow plot

[38]:

from src.visualisation.arrow import arrow_plot

arrow_plot(project="sdat2/ENSOTrend-gamma", show_plots=True, save_path=True)
arrow_plot(project="sdat2/ENSOTrend-beta", show_plots=True, save_path=True)
arrow_plot(project="sdat2/seager19", show_plots=True, save_path=True)

'summary_table'  1.88228 s

'_aggregate_matches'  0.00849 s

'summary_table'  1.96762 s

'_aggregate_matches'  0.00844 s

0.411
0 EECE 0.06+/-0.06 green
0 EEEC 0.32+/-0.11 orange
0 EECC 0.50+/-0.14 red
[1, 2, 3] ['W\n 0.06±0.06 K', 'RH\n 0.3±0.1 K', 'RH+W\n 0.5±0.1 K']
Difference 0.478
0.411
1 EESE -0.00+/-0.05 green
1 EEES 0.13+/-0.07 orange
1 EESS 0.14+/-0.07 red
[1, 2, 3] ['W\n -0.00±0.05 K', 'RH\n 0.1±0.1 K', 'RH+W\n 0.1±0.1 K']
Difference 0.361

<Figure size 3307.09x2043.89 with 0 Axes>
../_images/notebooks_analyse_results_33_2.png 
'summary_table'  1.20721 s

'_aggregate_matches'  0.00866 s

'summary_table'  2.16314 s

'_aggregate_matches'  0.00832 s

0.411
0 EECE 0.05+/-0.07 green
0 EEEC 0.29+/-0.12 orange
0 EECC 0.45+/-0.14 red
[1, 2, 3] ['W\n 0.05±0.07 K', 'RH\n 0.3±0.1 K', 'RH+W\n 0.5±0.1 K']
Difference 0.478
0.411
1 EESE -0.01+/-0.06 green
1 EEES 0.13+/-0.08 orange
1 EESS 0.18+/-0.05 red
[1, 2, 3] ['W\n -0.01±0.06 K', 'RH\n 0.1±0.1 K', 'RH+W\n 0.2±0.0 K']
Difference 0.361

<Figure size 3307.09x2043.89 with 0 Axes>
../_images/notebooks_analyse_results_33_5.png 
'summary_table'  3.74761 s

'_aggregate_matches'  0.00919 s

'summary_table'  5.41381 s

'_aggregate_matches'  0.00867 s

0.411
0 EECE 0.06+/-0.08 green
0 EEEC 0.34+/-0.09 orange
0 EECC 0.45+/-0.14 red
[1, 2, 3] ['W\n 0.06±0.08 K', 'RH\n 0.3±0.1 K', 'RH+W\n 0.5±0.1 K']
Difference 0.478
0.411
1 EESE 0.06+/-0.07 green
1 EEES 0.14+/-0.08 orange
1 EESS 0.28+/-0.10 red
[1, 2, 3] ['W\n 0.06±0.07 K', 'RH\n 0.1±0.1 K', 'RH+W\n 0.3±0.1 K']
Difference 0.361

<Figure size 3307.09x2043.89 with 0 Axes>
../_images/notebooks_analyse_results_33_8.png

Multi bias plot

[7]:

from src.visualisation.bias import multi_bias_plot
print("CMIP6")
multi_bias_plot("S")
plt.show()
print("CMIP5")
multi_bias_plot("C")
plt.show()

CMIP6
      CMIP5 MMM  LENS MMM  NCEP NCAR   ERSSTv5   HadlSST
2008   0.594690  0.460177        NaN  0.527434  0.265487
2009   0.614159  0.499115   0.238938  0.495575  0.267257
2010   0.647788  0.536283   0.109735  0.355752  0.134513
2011   0.681416  0.541593   0.015929  0.306195  0.077876
2012   0.702655  0.557522   0.017699  0.299115  0.072566
2013   0.752212  0.623009   0.021239  0.334513  0.097345
2014   0.785841  0.679646   0.010619  0.295575  0.079646
2015   0.810619  0.670796   0.086726  0.330973  0.118584
2016   0.838938  0.667257   0.077876  0.293805  0.076106
2017   0.872566  0.704425   0.129204  0.353982  0.106195
'plot_nino_box'  0.00041 s

'plot_nino_box'  0.00037 s

'plot_nino_box'  0.00036 s


<Figure size 3307.09x2043.89 with 0 Axes>
../_images/notebooks_analyse_results_35_2.png 
CMIP5
      CMIP5 MMM  LENS MMM  NCEP NCAR   ERSSTv5   HadlSST
2008   0.594690  0.460177        NaN  0.527434  0.265487
2009   0.614159  0.499115   0.238938  0.495575  0.267257
2010   0.647788  0.536283   0.109735  0.355752  0.134513
2011   0.681416  0.541593   0.015929  0.306195  0.077876
2012   0.702655  0.557522   0.017699  0.299115  0.072566
2013   0.752212  0.623009   0.021239  0.334513  0.097345
2014   0.785841  0.679646   0.010619  0.295575  0.079646
2015   0.810619  0.670796   0.086726  0.330973  0.118584
2016   0.838938  0.667257   0.077876  0.293805  0.076106
2017   0.872566  0.704425   0.129204  0.353982  0.106195
[Errno 2] No such file or directory: b'/home/users/sithom/seager19/atmos/DATA/ts-CMIP5-trend.nc'
'plot_nino_box'  0.00036 s

'plot_nino_box'  0.00035 s


<Figure size 3307.09x2043.89 with 0 Axes>
../_images/notebooks_analyse_results_35_5.png