Depth sweep#

Compare performance as mesh resolution increases. Generates ch3-4 data.

import xcell as xc
import Common_nongallery
import matplotlib.pyplot as plt
import argparse
import numpy as np

cli = argparse.ArgumentParser()
cli.add_argument("--comparison", choices=["bounds", "mesh", "formula", "bigPOC", "fixedDisc"], default="fixedDisc")
cli.add_argument("-p", "--plot-only", help="skip simulation and use existing data", action="store_true")
# cli.add_argument('-a','--animate',help='skip simulation and use existing data', action = 'store_true')
# cli.add_argument('-p','--plot-only',help='skip simulation and use existing data', action = 'store_true')

args = cli.parse_args()


generate = True

# plot performance info
staticPlots = True

depths = np.arange(3, 12)

xtraParams = None
xmax = 1e-4
if args.comparison == "mesh" or args.comparison == "bigPOC":
    foldername = "Comparison/PoC"
    tstVals = ["adaptive", "uniform"]
    # tstVals=['adaptive','equal elements',r'equal $l_0$']
    tstCat = "Mesh type"
if args.comparison == "formula" or args.comparison == "fixedDisc":
    foldername = "Comparison/formulations"
    tstVals = ["Admittance", "FEM", "Face"]
    tstCat = "Element type"
if args.comparison == "bounds":
    foldername = "Comparison/boundaries"
    tstVals = ["Analytic", "Ground", "Rubik0"]
    tstCat = "Boundary"
if args.comparison == "testing":
    foldername = "Comparison/miniset"
    tstVals = ["adaptive", "uniform"]
    tstCat = "Mesh type"
    generate = False
    staticPlots = False
    depths = np.arange(3, 8)

if args.comparison == "bigPOC":
    foldername = "Comparison/bigPOC"
    xmax = 1e-2

if args.comparison == "fixedDisc":
    foldername = "Comparison/fixedDisc"
    xtraParams = {"boundary_functionction": "Analytic"}


# if args.comparison=='voltage':
# tstVals=[False, True]
# tstCat='Vsrc?'


# generate animation(s)
plotters = [
    xc.visualizers.ErrorGraph,
    # xc.visualizers.ErrorGraph,
    # xc.visualizers.SliceSet,
    # xc.visualizers.LogError,
    # xc.visualizers.CurrentPlot,
]

plotPrefs = [
    None,
    # {'onlyDoF':True},
    # None,
    # None,
]


study, _ = Common_nongallery.makeSynthStudy(foldername, xmax=xmax)

if generate and not args.plot_only:
    Common_nongallery.pairedDepthSweep(
        study, depthRange=depths, testCat=tstCat, testVals=tstVals, overrides=xtraParams
    )

2.58777 Gb used
error: 1.39509
2.58777 Gb used
error: 1.38292

Calculating adjacency:   0%|          | 0/120 [00:00<?, ?it/s]
Calculating adjacency:   1%|          | 1/120 [00:00<00:29,  3.99it/s]
Calculating adjacency: 100%|██████████| 120/120 [00:00<00:00, 475.24it/s]
/Users/cgirard/mambaforge/envs/xcell/lib/python3.9/site-packages/scipy/sparse/linalg/_dsolve/linsolve.py:206: MatrixRankWarning: Matrix is exactly singular
  warn("Matrix is exactly singular", MatrixRankWarning)
2.5873 Gb used
error: nan
2.58787 Gb used
error: 0.711183
2.58792 Gb used
error: 0.626179

Calculating adjacency:   0%|          | 0/512 [00:00<?, ?it/s]
Calculating adjacency: 100%|██████████| 512/512 [00:00<00:00, 67291.81it/s]
/Users/cgirard/mambaforge/envs/xcell/lib/python3.9/site-packages/scipy/sparse/linalg/_dsolve/linsolve.py:206: MatrixRankWarning: Matrix is exactly singular
  warn("Matrix is exactly singular", MatrixRankWarning)
2.58792 Gb used
error: nan
2.58795 Gb used
error: 0.358631
2.58805 Gb used
error: 0.350071

Calculating adjacency:   0%|          | 0/736 [00:00<?, ?it/s]
Calculating adjacency: 100%|██████████| 736/736 [00:00<00:00, 65509.58it/s]
2.58807 Gb used
error: nan
2.58807 Gb used
error: 0.203653
2.58816 Gb used
error: 0.504739

Calculating adjacency:   0%|          | 0/1632 [00:00<?, ?it/s]
Calculating adjacency: 100%|██████████| 1632/1632 [00:00<00:00, 64656.97it/s]
2.58818 Gb used
error: nan
2.58818 Gb used
error: 0.298805
2.5882 Gb used
error: 0.596175

Calculating adjacency:   0%|          | 0/2976 [00:00<?, ?it/s]
Calculating adjacency: 100%|██████████| 2976/2976 [00:00<00:00, 60473.96it/s]
2.58839 Gb used
error: nan
2.58944 Gb used
error: 0.248566
2.60198 Gb used
error: 0.429459

Calculating adjacency:   0%|          | 0/5272 [00:00<?, ?it/s]
Calculating adjacency:  80%|███████▉  | 4211/5272 [00:00<00:00, 12583.86it/s]
Calculating adjacency: 100%|██████████| 5272/5272 [00:00<00:00, 14963.84it/s]
2.60207 Gb used
error: nan
2.60207 Gb used
error: 0.239598
2.63931 Gb used
error: 0.415435

Calculating adjacency:   0%|          | 0/7232 [00:00<?, ?it/s]
Calculating adjacency:  79%|███████▉  | 5721/7232 [00:00<00:00, 57198.46it/s]
Calculating adjacency: 100%|██████████| 7232/7232 [00:00<00:00, 57282.29it/s]
2.64092 Gb used
error: nan
2.64135 Gb used
error: 0.233044
2.67048 Gb used
error: 0.368771

Calculating adjacency:   0%|          | 0/14232 [00:00<?, ?it/s]
Calculating adjacency:  38%|███▊      | 5341/14232 [00:00<00:00, 53404.58it/s]
Calculating adjacency:  75%|███████▌  | 10698/14232 [00:00<00:00, 53498.46it/s]
Calculating adjacency: 100%|██████████| 14232/14232 [00:00<00:00, 53738.71it/s]
2.77851 Gb used
error: nan
2.79457 Gb used
error: 0.226043
3.29538 Gb used
error: 0.334305

Calculating adjacency:   0%|          | 0/24536 [00:00<?, ?it/s]
Calculating adjacency:  21%|██▏       | 5227/24536 [00:00<00:00, 52268.43it/s]
Calculating adjacency:  43%|████▎     | 10454/24536 [00:00<00:00, 52060.13it/s]
Calculating adjacency:  64%|██████▍   | 15704/24536 [00:00<00:00, 52256.48it/s]
Calculating adjacency:  85%|████████▌ | 20930/24536 [00:00<00:00, 51626.94it/s]
Calculating adjacency: 100%|██████████| 24536/24536 [00:00<00:00, 52150.34it/s]
3.46415 Gb used
error: nan

costcat = "Error"
# costcat='FVU'
# x_category='l0min'

xvalues = ["Number of elements", "min_l0", "Total time [Wall]"]
xtags = ["numel", "l0", "totTime"]
if staticPlots:
    for x_category, xtag in zip(xvalues, xtags):
        xc.visualizers.grouped_scatter(
            study.study_path + "/log.csv", x_category=x_category, y_category=costcat, group_category=tstCat
        )
        fname = tstCat + "_" + costcat + "-vs-" + xtag
        fname.replace(" ", "_")
        nufig = plt.gcf()
        study.save_plot(nufig, fname)
        for fv in tstVals:
            fstack, fratio = xc.visualizers.plot_study_performance(
                study, plot_ratios=True, only_category=tstCat, only_value=fv
            )
            fstem = "_" + tstCat + str(fv)

            study.save_plot(fstack, "Performance" + fstem)

            study.save_plot(fratio, "Ratio" + fstem)

/Users/cgirard/repos/xcell/xcell/visualizers.py:437: RuntimeWarning: More than 20 figures have been opened. Figures created through the pyplot interface (`matplotlib.pyplot.figure`) are retained until explicitly closed and may consume too much memory. (To control this warning, see the rcParam `figure.max_open_warning`).
  fig = plt.figure()

for ii, p in enumerate(plotters):
    plots = []
    names = []
    ranges = None
    for fv in tstVals:
        fname = p.__name__ + "_" + str(fv)
        fname.replace(" ", "_")
        plotr = p(plt.figure(), study, prefs=plotPrefs[ii])
        if "universalPts" in plotr.prefs:
            plotr.prefs["universalPts"] = True
        if "onlyDoF" in plotr.prefs:
            if plotr.prefs["onlyDoF"]:
                fname += "-detail"

        plotr.get_study_data(filter_categories=[tstCat], filter_values=[fv])

        plots.append(plotr)
        names.append(fname)

        if ranges is not None:
            plotr.unify_scales(ranges)
        ranges = plotr.data_scales

    for plot, name in zip(plots, names):
        plot.data_scales = ranges

        plot.animate_study(fname=name, fps=1.0)

Total running time of the script: (3 minutes 4.150 seconds)

Gallery generated by Sphinx-Gallery