def load_reshape_dataset(fname):
"""
Load dataset as OrderedDict. Each element of the
OrderedDict is a multidimensional numpy array, in
a way that getting the result for a specific
composition becomes very easy.
"""
# read database header to get data structure
header = read_header_database(fname)
# temperature range; composition range
Trange, crange = parse_header_database(header)
# read database
df = load_dataset(fname)
# index
df['idx'] = np.array([get_idx_from_fname(fname)
for fname in df['file']])
# get shape of multidimensional array
newshape = []
for el, rng in crange.items():
newshape.append(rng.lvls)
# reshape flattened dataframe as dictionary of multidimensional arrays
dataset = OrderedDict()
for key in df.columns:
dataset[key] = df[key].values.reshape(newshape)
return dataset
import Tcritical_regression_statsmodels as tr
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
from scipy.optimize import root
from parse_database import read_header_database, parse_header_database
from plot_carbon_isopleth import (load_reshape_dataset, select_carbon_isopleth,
plot_carbon_isopleth)
fname_dataset = '../databases/Tcritical.csv'
header = read_header_database(fname_dataset)
# Temperature range and compositions ranges
Trange, crange = parse_header_database(header)
# dataset as pandas DataFrame
dataset = tr.load_dataset(fname_dataset)
# dependent variable
dep_var = 'A1'
reg, results = {}, {}
reg, results = tr.regression_poly_2nd_deg(dataset=dataset,
dep_var=dep_var,
maxpvalue=.01,
printsummary=False)
ax = tr.plot_fitting_results(results, dataset, dep_var)
ax.set_title('{}'.format(dep_var))