Closed tkoskela closed 1 year ago
Currently plotting the following gives an error because we force all series to contain all x axis values:
x_axis: value: "test_name" units: custom: null y_axis: value: "flops_value" units: column: "flops_unit" filters: [] series: [["system", "archer2"], ["system", "isambard-macs"]]
RuntimeError: ('Unexpected number of rows (9) does not match number of unique x-axis values per series (12)', test_name flops_value flops_unit system 0 HPCG_IntelSKX_Isambard 38.6863 Gflops/seconds isambard-macs 1 HPCG_Original 24.0270 Gflops/seconds isambard-macs 2 HPCG_Stencil 51.0070 Gflops/seconds isambard-macs 3 HPCG_IntelAVX2_Isambard 39.0427 Gflops/seconds isambard-macs 4 HPCG_LFRic 18.5399 Gflops/seconds isambard-macs 5 HPCG_IntelAVX_Isambard 38.7652 Gflops/seconds isambard-macs 6 HPCG_Original 39.2325 Gflops/seconds archer2 7 HPCG_Stencil 124.1970 Gflops/seconds archer2 8 HPCG_LFRic 55.9839 Gflops/seconds archer2) Post-processing stopped
We should be able to relax this to allowing the total number of values to be less or equal the number of series * the number of x axis values
Currently plotting the following gives an error because we force all series to contain all x axis values:
We should be able to relax this to allowing the total number of values to be less or equal the number of series * the number of x axis values