import random from colornames.convex_hull_classifier import ConvexHullClassifier from colornames.diagnostics import write_diagnostic_html, random_cie_colors, make_cie_gradient import common coords, names = common.read_old_data("../../data/old_data.csv") thresholds = [2.0, 0.0, 0.0] clf = ConvexHullClassifier(thresholds) clf.fit(coords, names) y, dist, thresh, ynames, inhull, numhulls = clf.predict(coords) write_diagnostic_html("../color_page.html", coords, names, y, ynames, dist, thresh, inhull, numhulls) random.seed(123456) rcie = random_cie_colors(200) yr, distr, threshr, ynamesr, inhullr, numhullsr = clf.predict(rcie) write_diagnostic_html("../color_rcie.html", rcie, None, yr, ynamesr, distr, threshr, inhullr, numhullsr) gcie = make_cie_gradient(200, clf.hull_centroid("RED"), clf.hull_centroid("ORANGE")) yg, distg, threshg, ynamesg, inhullg, numhullsg = clf.predict(gcie) write_diagnostic_html("../color_gcie.html", gcie, None, yg, ynamesg, distg, threshg, inhullg, numhullsg)
import pandas as pd from colornames.convex_hull_classifier import ConvexHullClassifier from colornames.diagnostics import write_diagnostic_html from common import read_old_data coords, names = read_old_data('../../data/old_data.csv') thresholds = [2.0, 0.0, 0.0] clf = ConvexHullClassifier(thresholds) clf.fit(coords, names) new_data_filename = '../../data/new_data.csv' new_data = pd.read_csv(new_data_filename) new_coords = new_data.ix[:, 2:5] y, dist, thresh, ynames, inhull, numhulls = clf.predict(new_coords) dist_sort = dist.values dist_sort.sort(axis=1) dist_sort = pd.DataFrame(dist_sort, dist.index) dist_sort.columns = ['dist_sort_%s' % c for c in dist_sort.columns] dist0 = dist_sort.iloc[:, 0].copy() dist1 = dist_sort.iloc[:, 1].copy() selected = ( (numhulls < 1) | (numhulls > 1) | ((ynames == 'NEUTRAL') & (dist0 < 1.0) & (dist1 < 10.0)) | ((ynames != 'NEUTRAL') & (dist0 < 3.0) & (dist1 < 10.0)))
import pandas as pd from colornames.convex_hull_classifier import ConvexHullClassifier from colornames.diagnostics import write_diagnostic_html from common import read_old_data coords, names = read_old_data('../../data/old_data.csv') thresholds = [2.0, 0.0, 0.0] clf = ConvexHullClassifier(thresholds) clf.fit(coords, names) new_data_filename = '../../data/new_data.csv' new_data = pd.read_csv(new_data_filename) new_coords = new_data.ix[:, 2:5] y, dist, thresh, ynames, inhull, numhulls = clf.predict(new_coords) dist_sort = dist.values dist_sort.sort(axis=1) dist_sort = pd.DataFrame(dist_sort, dist.index) dist_sort.columns = ['dist_sort_%s' % c for c in dist_sort.columns] dist0 = dist_sort.iloc[:, 0].copy() dist1 = dist_sort.iloc[:, 1].copy() selected = ((numhulls < 1) | (numhulls > 1) | ((ynames == 'NEUTRAL') & (dist0 < 1.0) & (dist1 < 10.0)) | ((ynames != 'NEUTRAL') & (dist0 < 3.0) & (dist1 < 10.0))) write_diagnostic_html('../color_new.html', new_coords, None, y, ynames, dist, thresh, inhull, numhulls, selected)
import random from colornames.convex_hull_classifier import ConvexHullClassifier from colornames.diagnostics import write_diagnostic_html, random_cie_colors, make_cie_gradient import common coords, names = common.read_old_data('../../data/old_data.csv') thresholds = [2.0, 0.0, 0.0] clf = ConvexHullClassifier(thresholds) clf.fit(coords, names) y, dist, thresh, ynames, inhull, numhulls = clf.predict(coords) write_diagnostic_html('../color_page.html', coords, names, y, ynames, dist, thresh, inhull, numhulls) random.seed(123456) rcie = random_cie_colors(200) yr, distr, threshr, ynamesr, inhullr, numhullsr = clf.predict(rcie) write_diagnostic_html('../color_rcie.html', rcie, None, yr, ynamesr, distr, threshr, inhullr, numhullsr) gcie = make_cie_gradient(200, clf.hull_centroid('RED'), clf.hull_centroid('ORANGE')) yg, distg, threshg, ynamesg, inhullg, numhullsg = clf.predict(gcie) write_diagnostic_html('../color_gcie.html', gcie, None, yg, ynamesg, distg, threshg, inhullg, numhullsg)