def plot_swe(): seasons = ["(a) Annual", " (b) Winter (DJF)", "(c) Spring (MAM)", "(d) Summer (JJA)", "(e) Fall (SON)"] months = [range(1, 13), [12, 1, 2], [3, 4, 5], [6, 7, 8], [9, 10, 11] ] season_to_months = dict(zip(seasons, months)) var_name = "sno" year_range_c = xrange(1970,2000) year_range_f = xrange(2041,2071) x, y = polar_stereographic.xs, polar_stereographic.ys i_array, j_array = _get_routing_indices() x_min, x_max, y_min, y_max = plot_utils.get_ranges(x[i_array, j_array], y[i_array, j_array]) _generate_mask_of_domain_of_interest(i_array, j_array) if True: return plot_utils.apply_plot_params(width_pt= None, font_size=9, aspect_ratio=2.5) fig = plt.figure() assert isinstance(fig, Figure) basemap = polar_stereographic.basemap assert isinstance(basemap, Basemap) gs = gridspec.GridSpec(3,4, height_ratios=[1,1,1], width_ratios=[1,1,1,1]) #color_map = my_cm.get_ color_map = my_cm.get_red_blue_colormap(ncolors = 10, reversed=True) #color_map = mpl.cm.get_cmap(name="jet_r", lut=10) clevels = xrange(-8, 9, 2) all_plot_axes = [] for i, season in enumerate(seasons): if not i: ax = fig.add_subplot(gs[0,1:3]) else: row, col = (i - 1) // 2 + 1, (i - 1) % 2 ax = fig.add_subplot(gs[row, col * 2 : col * 2 + 2 ]) all_plot_axes.append(ax) current = _get_data(v_name=var_name, months = season_to_months[season], member_list=members.current_ids, year_range=year_range_c) future = _get_data(v_name=var_name, months = season_to_months[season], member_list=members.future_ids, year_range=year_range_f) t, p = stats.ttest_ind(current, future, axis=0) #TODO: change it back to p <= 0.05 wheen doing real sign test significant = np.array(p <= 1) assert not np.all(~significant) assert not np.all(significant) current_m = np.mean(current, axis=0) future_m = np.mean(future, axis= 0) delta = (future_m - current_m) delta = np.array(delta) assert isinstance(ax, Axes) #delta = maskoceans(polar_stereographic.lons, polar_stereographic.lats, delta) save = delta[i_array, j_array] delta = np.ma.masked_all(delta.shape) delta[i_array, j_array] = save d_min = np.floor( np.min(save) ) d_max = np.ceil( np.max(save) ) bounds = plot_utils.get_boundaries_for_colobar(d_min, d_max, color_map.N, lambda x: np.round(x, decimals=10)) print bounds bn = BoundaryNorm(bounds, color_map.N) d = np.max( np.abs([d_min, d_max]) ) print season, np.min(delta), np.max(delta) #delta = np.ma.masked_where(delta < 0, delta ) img = basemap.pcolormesh(x, y, delta, cmap = color_map, norm = bn, vmin = bounds[0], vmax = bounds[-1]) divider = make_axes_locatable(ax) cax = divider.append_axes("right", "8%", pad="3%") assert isinstance(cax, Axes) int_ticker = LinearLocator(numticks = color_map.N + 1) cb = fig.colorbar(img, cax = cax, ticks = bounds, boundaries = bounds) where_significant = significant significant = np.ma.masked_all(significant.shape) significant[(~where_significant)] = 0 save = significant[i_array, j_array] significant = np.ma.masked_all(significant.shape) significant[i_array, j_array] = save basemap.pcolormesh( x, y, significant , cmap = mpl.cm.get_cmap(name = "gray", lut = 3), vmin = -1, vmax = 1, ax = ax) ax.set_title(season) for the_ax in all_plot_axes: the_ax.set_xlim(x_min, x_max) the_ax.set_ylim(y_min, y_max) basemap.drawcoastlines(ax = the_ax, linewidth = 0.1) plot_utils.draw_meridians_and_parallels(basemap, step_degrees=30.0, ax = the_ax) plot_basin_boundaries_from_shape(basemap, axes = the_ax, linewidth=0.4) gs.update(wspace=0.5) #gs.tight_layout(fig) fig.suptitle("Projected changes, SWE, mm, CRCM4") fig.savefig("proj_change_{0}_ccc.png".format(var_name)) pass
def plot_temp(): seasons = ["(a) Annual", " (b) Winter (DJF)", "(c) Spring (MAM)", "(d) Summer (JJA)", "(e) Fall (SON)"] months = [range(1, 13), [12, 1, 2], [3, 4, 5], [6, 7, 8], [9, 10, 11] ] season_to_months = dict(zip(seasons, months)) #take out annual seasons.pop(0) #put new numbering for the subplots new_numbering = ["a", "b", "c", "d"] var_name = "st" year_range_c = xrange(1970,2000) year_range_f = xrange(2041,2071) x, y = polar_stereographic.xs, polar_stereographic.ys i_array, j_array = _get_routing_indices() x_min, x_max, y_min, y_max = plot_utils.get_ranges(x[i_array, j_array], y[i_array, j_array]) plot_utils.apply_plot_params(width_pt= None, font_size=9, aspect_ratio=2.5) fig = plt.figure() assert isinstance(fig, Figure) basemap = polar_stereographic.basemap assert isinstance(basemap, Basemap) gs = gridspec.GridSpec(3,2) #color_map = mpl.cm.get_cmap(name="jet", lut=10) clevels = xrange(-8, 9, 2) all_plot_axes = [] #determine min an max for color scales min_val = np.inf max_val = -np.inf for season in seasons: current = _get_data(v_name=var_name, months = season_to_months[season], member_list=members.current_ids, year_range=year_range_c) future = _get_data(v_name=var_name, months = season_to_months[season], member_list=members.future_ids, year_range=year_range_f) current_m = np.mean(current, axis=0) future_m = np.mean(future, axis= 0) delta = future_m[i_array, j_array] - current_m[i_array, j_array] the_min = delta.min() the_max = delta.max() min_val = min(min_val, the_min) max_val = max(max_val, the_max) min_val = np.floor(min_val) max_val = np.ceil(max_val) color_map = my_cm.get_red_blue_colormap(ncolors = 10, reversed=True) if min_val >= 0: color_map = my_cm.get_red_colormap(ncolors=10) for i, season in enumerate(seasons): row, col = i // 2, i % 2 ax = fig.add_subplot(gs[row, col ]) all_plot_axes.append(ax) current = _get_data(v_name=var_name, months = season_to_months[season], member_list=members.current_ids, year_range=year_range_c) future = _get_data(v_name=var_name, months = season_to_months[season], member_list=members.future_ids, year_range=year_range_f) # t, p = stats.ttest_ind(current, future, axis=0) # significant = np.array(p <= 0.05) significant = calculate_significance_using_bootstrap(current, future) assert not np.all(~significant) assert not np.all(significant) current_m = np.mean(current, axis=0) future_m = np.mean(future, axis= 0) delta = (future_m - current_m) assert isinstance(ax, Axes) #delta = maskoceans(polar_stereographic.lons, polar_stereographic.lats, delta) save = delta[i_array, j_array] delta = np.ma.masked_all(delta.shape) delta[i_array, j_array] = save d_min = np.floor( np.min(save) ) d_max = np.ceil( np.max(save) ) img = basemap.pcolormesh(x, y, delta, cmap = color_map, vmin = min_val, vmax = max_val) divider = make_axes_locatable(ax) cax = divider.append_axes("right", "8%", pad="3%") assert isinstance(cax, Axes) int_ticker = LinearLocator(numticks = color_map.N + 1) cb = fig.colorbar(img, cax = cax, ticks = int_ticker) cax.set_title("$^{\\circ}{\\rm C}$") where_significant = significant significant = np.ma.masked_all(significant.shape) significant[~where_significant] = 0 basemap.pcolormesh( x, y, significant , cmap = mpl.cm.get_cmap(name = "gray", lut = 3), vmin = -1, vmax = 1, ax = ax) ax.set_title( season.replace( re.findall( "([a-z])", season)[0], new_numbering[i]) ) for the_ax in all_plot_axes: the_ax.set_xlim(x_min, x_max) the_ax.set_ylim(y_min, y_max) basemap.drawcoastlines(ax = the_ax, linewidth = 0.1) plot_utils.draw_meridians_and_parallels(basemap, step_degrees=30.0, ax = the_ax) plot_basin_boundaries_from_shape(basemap, axes = the_ax, linewidth=0.4) #gs.update(wspace=0.5) fig.tight_layout() #fig.suptitle("Projected changes, T(2m), degrees, CRCM4") fig.savefig("proj_change_{0}_ccc.png".format(var_name))
def plot_precip(data_path = "/home/huziy/skynet1_rech3/crcm4_data"): seasons = ["(a) Annual", " (b) Winter (DJF)", "(c) Spring (MAM)", "(d) Summer (JJA)", "(e) Fall (SON)"] months = [range(1, 13), [12, 1, 2], [3, 4, 5], [6, 7, 8], [9, 10, 11] ] season_to_months = dict(zip(seasons, months)) #remove annual seasons.pop(0) #put new numbering for the subplots new_numbering = ["a", "b", "c", "d"] var_name = "pcp" year_range_c = xrange(1970,2000) year_range_f = xrange(2041,2071) x, y = polar_stereographic.xs, polar_stereographic.ys i_array, j_array = _get_routing_indices() x_min, x_max, y_min, y_max = plot_utils.get_ranges(x[i_array, j_array], y[i_array, j_array]) plot_utils.apply_plot_params(width_pt= None, font_size=9, aspect_ratio=2.5) fig = plt.figure() assert isinstance(fig, Figure) basemap = polar_stereographic.basemap assert isinstance(basemap, Basemap) gs = gridspec.GridSpec(3,2, height_ratios=[1,1,1], width_ratios=[1,1]) #determine min an max for color scales min_val = np.inf max_val = -np.inf for season in seasons: current = _get_data(v_name=var_name, months = season_to_months[season], member_list=members.current_ids, year_range=year_range_c) future = _get_data(v_name=var_name, months = season_to_months[season], member_list=members.future_ids, year_range=year_range_f) current_m = np.mean(current, axis=0) future_m = np.mean(future, axis= 0) delta = future_m[i_array, j_array] - current_m[i_array, j_array] the_min = delta.min() the_max = delta.max() min_val = min(min_val, the_min) max_val = max(max_val, the_max) min_val = np.floor(min_val) max_val = np.ceil(max_val) color_map = my_cm.get_red_blue_colormap(ncolors = 10, reversed=False) #color_map = mpl.cm.get_cmap(name="jet_r", lut=10) clevels = xrange(-8, 9, 2) all_plot_axes = [] for i, season in enumerate(seasons): row, col = i // 2, i % 2 ax = fig.add_subplot(gs[row, col ]) all_plot_axes.append(ax) current = _get_data(data_folder=data_path, v_name=var_name, months = season_to_months[season], member_list=members.current_ids, year_range=year_range_c) future = _get_data(data_folder=data_path, v_name=var_name, months = season_to_months[season], member_list=members.future_ids, year_range=year_range_f) #t, p = stats.ttest_ind(current, future, axis=0) #significant = np.array(p <= 0.05) significant = calculate_significance_using_bootstrap(current, future) assert not np.all(~significant) assert not np.all(significant) current_m = np.mean(current, axis=0) future_m = np.mean(future, axis= 0) seconds_per_day = 24 * 60 * 60 delta = (future_m - current_m) * seconds_per_day delta = np.array(delta) assert isinstance(ax, Axes) #delta = maskoceans(polar_stereographic.lons, polar_stereographic.lats, delta) save = delta[i_array, j_array] delta = np.ma.masked_all(delta.shape) delta[i_array, j_array] = save d_min = np.floor( min_val * 10 ) / 10.0 d_max = np.ceil( max_val *10 ) / 10.0 if d_min > 0: color_map = my_cm.get_blue_colormap(ncolors=10) img = basemap.pcolormesh(x, y, delta, cmap = color_map, vmin = d_min, vmax = d_max) divider = make_axes_locatable(ax) cax = divider.append_axes("right", "8%", pad="3%") assert isinstance(cax, Axes) int_ticker = LinearLocator(numticks = color_map.N + 1) cb = fig.colorbar(img, cax = cax, ticks = int_ticker) cax.set_title("mm/d") where_significant = significant significant = np.ma.masked_all(significant.shape) significant[(~where_significant)] = 0 save = significant[i_array, j_array] significant = np.ma.masked_all(significant.shape) significant[i_array, j_array] = save basemap.pcolormesh( x, y, significant , cmap = mpl.cm.get_cmap(name = "gray", lut = 3), vmin = -1, vmax = 1, ax = ax) ax.set_title( season.replace( re.findall( "([a-z])", season)[0], new_numbering[i]) ) #plot djf swe change season = " (b) Winter (DJF)" ax = fig.add_subplot(gs[2, : ]) all_plot_axes.append(ax) var_name = "sno" current = _get_data(data_folder=data_path, v_name=var_name, months = season_to_months[season], member_list=members.current_ids, year_range=year_range_c) future = _get_data(data_folder=data_path, v_name=var_name, months = season_to_months[season], member_list=members.future_ids, year_range=year_range_f) #t, p = stats.ttest_ind(current, future, axis=0) #significant = np.array(p <= 0.05) significant = calculate_significance_using_bootstrap(current, future) assert not np.all(~significant) assert not np.all(significant) current_m = np.mean(current, axis=0) future_m = np.mean(future, axis= 0) delta = future_m - current_m delta = np.array(delta) assert isinstance(ax, Axes) #delta = maskoceans(polar_stereographic.lons, polar_stereographic.lats, delta) save = delta[i_array, j_array] delta = np.ma.masked_all(delta.shape) delta[i_array, j_array] = save d_min = np.floor( np.min(save) * 10 ) / 10.0 d_max = np.ceil( np.max(save) *10 ) / 10.0 if d_min >= 0: color_map = my_cm.get_blue_colormap(ncolors=10) bounds = plot_utils.get_boundaries_for_colobar(d_min, d_max, color_map.N, lambda x: np.round(x, decimals=0)) bn = BoundaryNorm(bounds, color_map.N) img = basemap.pcolormesh(x, y, delta, cmap = color_map, vmin = bounds[0], vmax = bounds[-1], norm = bn) divider = make_axes_locatable(ax) cax = divider.append_axes("right", "8%", pad="3%") assert isinstance(cax, Axes) int_ticker = LinearLocator(numticks = color_map.N + 1) cb = fig.colorbar(img, cax = cax, ticks = bounds) cax.set_title("mm") where_significant = significant significant = np.ma.masked_all(significant.shape) significant[(~where_significant)] = 0 save = significant[i_array, j_array] significant = np.ma.masked_all(significant.shape) significant[i_array, j_array] = save basemap.pcolormesh( x, y, significant , cmap = mpl.cm.get_cmap(name = "gray", lut = 3), vmin = -1, vmax = 1, ax = ax) ax.set_title( season.replace( re.findall( "([a-z])", season)[0], "e")) #finish swe djf for the_ax in all_plot_axes: the_ax.set_xlim(x_min, x_max) the_ax.set_ylim(y_min, y_max) basemap.drawcoastlines(ax = the_ax, linewidth = 0.1) plot_utils.draw_meridians_and_parallels(basemap, step_degrees=30.0, ax = the_ax) plot_basin_boundaries_from_shape(basemap, axes = the_ax, linewidth=0.4) #gs.update(wspace=0.5) #gs.tight_layout(fig) #fig.suptitle("Projected changes, total precip (mm/day), CRCM4") fig.tight_layout() fig.savefig("proj_change_{0}_ccc.png".format(var_name)) pass