What am I missing in this context?

[luxiant]

Hi. I'm trying a project to calibrate stochastic volatility in option data. Inside my code, I'm going to try nelder-mead method in order to get the parameters that make an optimum of mean square error. Here is the part of my source code.

//...
    let l_a = (second_implied_volatility-first_implied_volatility)/(second_moneyness-first_moneyness);
    let l_b = ((first_moneyness*second_implied_volatility)-(second_moneyness*first_implied_volatility))/(first_moneyness-second_moneyness);
    let r_a = (last_first_implied_volatility-last_second_implied_volatility)/(last_first_moneyness-last_second_moneyness);
    let r_b = ((last_second_moneyness*last_first_implied_volatility)-(last_first_moneyness*last_second_implied_volatility))/(last_second_moneyness-last_first_moneyness);

    let clustered_df_clone = clustered_df.clone();
    let svi_initial_sigma = clustered_df_clone.column("implied_volatility").unwrap().clone().min().unwrap() as f64;
    let svi_initial_a = &l_b+(&l_a*((&r_b-&l_b)/(&l_a-&r_a)));
    let svi_initial_b = (&r_a-&l_a)/2.0;
    let svi_initial_rho = (&r_a+l_a)/(&r_a-&l_a);
    let svi_initial_m = (&l_a*(&r_b-&l_b))/(&svi_initial_b*(&l_a-&r_a)*(&svi_initial_rho-1.0));

    struct StochasticVolatility {
        moneyness_column: Series,
        implied_volatility_column: Series,
    }

    impl CostFunction for StochasticVolatility {
        type Param = Vec<f64>;
        type Output = f64;
        
        fn cost(&self, x: &Self::Param) -> Result<Self::Output, Error> {
            let length = self.moneyness_column.len();
            
            let sigma = x[0];
            let a = x[1];
            let b = x[2];
            let rho = x[3];
            let m = x[4];

            let mut volatility_square_error_vec: Vec<f64> = Vec::new();
            for i in 0..length {
                let moneyness = match self.moneyness_column.get(i).unwrap() {
                    AnyValue::Float64(x) => Some(x),
                    _ => None,
                }.unwrap();
                let actual_adjusted_implied_volatility = match self.implied_volatility_column.get(i).unwrap() {
                    AnyValue::Float64(x) => Some(x),
                    _ => None,
                }.unwrap();
                let svi_component_one = &rho*(&moneyness-&m);
                let svi_component_two_subcomponent_one = f64::powf(&moneyness-&m, 2.0);
                let svi_component_two_subcomponent_two = f64::powf(sigma, 2.0);
                let svi_component_two = (svi_component_two_subcomponent_one+svi_component_two_subcomponent_two).sqrt();
                let svi = &a+(&b*(&svi_component_one+&svi_component_two));
                let volatility_error = f64::powf(actual_adjusted_implied_volatility-svi, 2.0);
                volatility_square_error_vec.push(volatility_error);
            }

            let volatility_error_series = Series::new("volatility_error", &volatility_square_error_vec);
            let volatility_mean_square_error = volatility_error_series.mean().unwrap();

            Ok(volatility_mean_square_error)
        }
    }

    let cost = StochasticVolatility {
        moneyness_column: clustered_df.column("moneyness").unwrap().clone(),
        implied_volatility_column: clustered_df.column("implied_volatility").unwrap().clone(),
    };
    
    let initial_guess = vec![vec![svi_initial_sigma, svi_initial_a, svi_initial_b, svi_initial_rho, svi_initial_m]];

    let solver = NelderMead::new(initial_guess)
    .with_sd_tolerance(1e-6).unwrap();

    let res = Executor::new(cost, solver)
    .configure(|state| state.max_iters(300))
    .run().unwrap();

    let calibrated_svi_params_option = res.state().get_best_param().copied().unwrap();
    let calibrated_svi_params = match calibrated_svi_params_option {
        Some(x) => x,
        None => return None,
    };
    let calibrated_sigma = calibrated_svi_params[0];
    let calibrated_a = calibrated_svi_params[1];
    let calibrated_b = calibrated_svi_params[2];
    let calibrated_rho = calibrated_svi_params[3];
    let calibrated_m = calibrated_svi_params[4];
// ...

In this code, when I define "solver", the compiler says "the trait bound Vec<f64>: argmin_math::ArgminAdd<Vec<f64>, Vec<f64>> is not satisfied". Since I'm a newbie in rust, I understood that some trait bounds are missing but I have no idea what are they. Can you guys help me fixing this error?

[stefan-k]

Hi, could you post how you included argmin and argmin-math in your Cargo.toml, please?

[luxiant]

Uhh...Sorry to bother you. I found that the root cause of the problem that I upload at the first thread lies at the prior process, so I took some days to deal with that. Now I believe that I debugged them. And then I modified the current part of the code like this.

    let l_a = (second_implied_volatility-first_implied_volatility)/(second_moneyness-first_moneyness);
    let l_b = ((first_moneyness*second_implied_volatility)-(second_moneyness*first_implied_volatility))/(first_moneyness-second_moneyness);
    let r_a = (last_first_implied_volatility-last_second_implied_volatility)/(last_first_moneyness-last_second_moneyness);
    let r_b = ((last_second_moneyness*last_first_implied_volatility)-(last_first_moneyness*last_second_implied_volatility))/(last_second_moneyness-last_first_moneyness);

    let moneyness_vol_df_clone = moneyness_vol_df.clone();
    let svi_initial_sigma = moneyness_vol_df_clone.column("adjusted_implied_volatility").unwrap().clone().min().unwrap();
    let svi_initial_a = &l_b+(&l_a*((&r_b-&l_b)/(&l_a-&r_a)));
    let svi_initial_b = (&r_a-&l_a)/2.0;
    let svi_initial_rho = (&r_a+l_a)/(&r_a-&l_a);
    let svi_initial_m = (&l_a*(&r_b-&l_b))/(&svi_initial_b*(&l_a-&r_a)*(&svi_initial_rho-1.0));
    // check if the initial parameters calculated properly. Delete them later.
    println!("svi_initial_sigma: {:?}", svi_initial_sigma);
    println!("svi_initial_a: {:?}", svi_initial_a);
    println!("svi_initial_b: {:?}", svi_initial_b);
    println!("svi_initial_rho: {:?}", svi_initial_rho);
    println!("svi_initial_m: {:?}", svi_initial_m);

    struct StochasticVolatility {
        moneyness_column: Series,
        implied_volatility_column: Series,
    }

    impl CostFunction for StochasticVolatility {
        type Param = Vec<f64>;
        type Output = f64;
        
        fn cost(&self, x: &Self::Param) -> Result<Self::Output, Error> {
            let length = self.moneyness_column.len();
            
            let sigma = x[0];
            let a = x[1];
            let b = x[2];
            let rho = x[3];
            let m = x[4];

            let mut volatility_square_error_vec: Vec<f64> = Vec::new();
            for i in 0..length {
                let moneyness = match self.moneyness_column.get(i).unwrap() {
                    AnyValue::Float64(x) => Some(x),
                    _ => None,
                }.unwrap();
                let actual_adjusted_implied_volatility = match self.implied_volatility_column.get(i).unwrap() {
                    AnyValue::Float64(x) => Some(x),
                    _ => None,
                }.unwrap();
                let svi_component_one = &rho*(&moneyness-&m);
                let svi_component_two_subcomponent_one = f64::powf(&moneyness-&m, 2.0);
                let svi_component_two_subcomponent_two = f64::powf(sigma, 2.0);
                let svi_component_two = (svi_component_two_subcomponent_one+svi_component_two_subcomponent_two).sqrt();
                let svi = &a+(&b*(&svi_component_one+&svi_component_two));
                let volatility_error = f64::powf(actual_adjusted_implied_volatility-svi, 2.0);
                volatility_square_error_vec.push(volatility_error);
            }

            let volatility_error_series = Series::new("volatility_error", &volatility_square_error_vec);
            let volatility_mean_square_error = volatility_error_series.mean().unwrap();

            Ok(volatility_mean_square_error)
        }
    }

    let cost = StochasticVolatility {
        moneyness_column: moneyness_vol_df.column("moneyness").unwrap().clone(),
        implied_volatility_column: moneyness_vol_df.column("adjusted_implied_volatility").unwrap().clone(),
    };
    // I'm not sure with setting initial guess.
    let initial_guess = vec![
        vec![svi_initial_sigma, svi_initial_a, svi_initial_b, svi_initial_rho, svi_initial_m],
        vec![svi_initial_sigma-1.0, svi_initial_a, svi_initial_b, svi_initial_rho, svi_initial_m],
        vec![svi_initial_sigma, svi_initial_a-1.0, svi_initial_b, svi_initial_rho, svi_initial_m],
        vec![svi_initial_sigma, svi_initial_a, svi_initial_b-1.0, svi_initial_rho, svi_initial_m],
        vec![svi_initial_sigma, svi_initial_a, svi_initial_b, svi_initial_rho-1.0, svi_initial_m],
        vec![svi_initial_sigma, svi_initial_a, svi_initial_b, svi_initial_rho, svi_initial_m-1.0],
    ];

    let solver = NelderMead::new(initial_guess)
    .with_sd_tolerance(1e-6).unwrap();

    let res = Executor::new(cost, solver)
    .configure(|state| state.max_iters(300))
    .run().unwrap();

    let calibrated_svi_params_option = res.state().get_best_param();
    let calibrated_svi_params = match calibrated_svi_params_option {
        Some(x) => x.clone(),
        None => return None,
    };
    let calibrated_sigma = calibrated_svi_params[0];
    let calibrated_a = calibrated_svi_params[1];
    let calibrated_b = calibrated_svi_params[2];
    let calibrated_rho = calibrated_svi_params[3];
    let calibrated_m = calibrated_svi_params[4];

And here is what I got as an error.

moneyness_vol_df: shape: (197, 2)
┌───────────┬─────────────────────────────┐
│ moneyness ┆ adjusted_implied_volatility │
│ ---       ┆ ---                         │
│ f64       ┆ f64                         │
╞═══════════╪═════════════════════════════╡
│ -0.236858 ┆ 0.842261                    │
│ -0.201767 ┆ 0.73169                     │
│ -0.183748 ┆ 0.673636                    │
│ -0.165399 ┆ 0.612655                    │
│ …         ┆ …                           │
│ 1.050996  ┆ 3.684781                    │
│ 1.184528  ┆ 4.023828                    │
│ 1.338678  ┆ 4.400856                    │
│ 1.521     ┆ 4.829287                    │
└───────────┴─────────────────────────────┘
svi_initial_sigma: 0.002699815627238849
svi_initial_a: 0.7599437789589143
svi_initial_b: 2.750417037986357
svi_initial_rho: -0.14563246244579822
svi_initial_m: -0.2107334004556779
thread 'main' panicked at /root/.cargo/registry/src/index.crates.io-6f17d22bba15001f/argmin-0.8.1/src/solver/neldermead/mod.rs:355:43:
attempt to subtract with overflow
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

I first thought that I set numbers of initial parameter vector less than the solver requires (which is 6 I guess), but it brought the same error even after I increased the number of vectors in "initial_guess".

What am I missing?

[stefan-k]

This is really odd ... I don't see anything that could cause this. Could you add an observer to the Executor with .add_observer(SlogLogger::term(), ObserverMode::Always) and show the output again, please?

[luxiant]

Sure. Here it is.

moneyness_vol_df: shape: (197, 2)
┌───────────┬─────────────────────────────┐
│ moneyness ┆ adjusted_implied_volatility │
│ ---       ┆ ---                         │
│ f64       ┆ f64                         │
╞═══════════╪═════════════════════════════╡
│ -0.236858 ┆ 0.842261                    │
│ -0.201767 ┆ 0.73169                     │
│ -0.183748 ┆ 0.673636                    │
│ -0.165399 ┆ 0.612655                    │
│ …         ┆ …                           │
│ 1.050996  ┆ 3.684781                    │
│ 1.184528  ┆ 4.023828                    │
│ 1.338678  ┆ 4.400856                    │
│ 1.521     ┆ 4.829287                    │
└───────────┴─────────────────────────────┘
svi_initial_sigma: 0.002699815627238849
svi_initial_a: 0.7599437789589143
svi_initial_b: 2.750417037986357
svi_initial_rho: -0.14563246244579822
svi_initial_m: -0.2107334004556779
Dec 13 16:05:48.473 INFO Nelder-Mead method, max_iters: 300
Dec 13 16:05:48.473 INFO iter: 0, cost: 0.09720966266236483, best_cost: 0.09720966266236483, cost_count: 7, action: Reflection, time: 0.00006916
Dec 13 16:05:48.473 INFO iter: 1, cost: 0.09720966266236483, best_cost: 0.09720966266236483, cost_count: 8, time: 0.000074028, action: Reflection
Dec 13 16:05:48.473 INFO iter: 2, cost: 0.09720966266236483, best_cost: 0.09720966266236483, cost_count: 10, time: 0.000136866, action: ContractionOutside
Dec 13 16:05:48.473 INFO iter: 3, cost: 0.09720966266236483, best_cost: 0.09720966266236483, cost_count: 12, action: ContractionInside, time: 0.000121908
Dec 13 16:05:48.474 INFO iter: 4, cost: 0.09720966266236483, best_cost: 0.09720966266236483, cost_count: 14, time: 0.000180668, action: ContractionOutside
Dec 13 16:05:48.474 INFO iter: 5, cost: 0.09720966266236483, best_cost: 0.09720966266236483, cost_count: 16, action: ContractionInside, time: 0.000185599
Dec 13 16:05:48.474 INFO iter: 6, cost: 0.09720966266236483, best_cost: 0.09720966266236483, cost_count: 17, time: 0.00009971, action: Reflection
Dec 13 16:05:48.474 INFO iter: 7, cost: 0.09720966266236483, best_cost: 0.09720966266236483, cost_count: 18, time: 0.000072617, action: Reflection
Dec 13 16:05:48.474 INFO iter: 8, cost: 0.09720966266236483, best_cost: 0.09720966266236483, cost_count: 20, time: 0.000124378, action: ContractionOutside
Dec 13 16:05:48.474 INFO iter: 9, cost: 0.09720966266236483, best_cost: 0.09720966266236483, cost_count: 21, time: 0.000060023, action: Reflection
Dec 13 16:05:48.474 INFO iter: 10, cost: 0.09720966266236483, best_cost: 0.09720966266236483, cost_count: 22, time: 0.000085034, action: Reflection
Dec 13 16:05:48.474 INFO iter: 11, cost: 0.09720966266236483, best_cost: 0.09720966266236483, cost_count: 24, time: 0.000164121, action: ContractionInside
Dec 13 16:05:48.475 INFO iter: 12, cost: 0.06976217276300072, best_cost: 0.06976217276300072, cost_count: 26, time: 0.000120472, action: Expansion
Dec 13 16:05:48.475 INFO iter: 13, cost: 0.06976217276300072, best_cost: 0.06976217276300072, cost_count: 28, time: 0.000167083, action: ContractionInside
Dec 13 16:05:48.475 INFO iter: 14, cost: 0.06976217276300072, best_cost: 0.06976217276300072, cost_count: 29, action: Reflection, time: 0.000105939
Dec 13 16:05:48.475 INFO iter: 15, cost: 0.06976217276300072, best_cost: 0.06976217276300072, cost_count: 31, action: ContractionInside, time: 0.000122518
Dec 13 16:05:48.475 INFO iter: 16, cost: 0.06976217276300072, best_cost: 0.06976217276300072, cost_count: 33, action: ContractionInside, time: 0.00016207
Dec 13 16:05:48.475 INFO iter: 17, cost: 0.06814235910679896, best_cost: 0.06814235910679896, cost_count: 35, action: ContractionInside, time: 0.000130043
Dec 13 16:05:48.475 INFO iter: 18, cost: 0.06814235910679896, best_cost: 0.06814235910679896, cost_count: 37, action: ContractionInside, time: 0.000117193
Dec 13 16:05:48.476 INFO iter: 19, cost: 0.05484327425649087, best_cost: 0.05484327425649087, cost_count: 39, time: 0.000150972, action: Expansion
Dec 13 16:05:48.476 INFO iter: 20, cost: 0.05484327425649087, best_cost: 0.05484327425649087, cost_count: 40, action: Reflection, time: 0.00010676
Dec 13 16:05:48.476 INFO iter: 21, cost: 0.05484327425649087, best_cost: 0.05484327425649087, cost_count: 41, action: Reflection, time: 0.00009601
Dec 13 16:05:48.476 INFO iter: 22, cost: 0.050129923661369334, best_cost: 0.050129923661369334, cost_count: 43, action: ContractionInside, time: 0.000155743
Dec 13 16:05:48.476 INFO iter: 23, cost: 0.04981798346803804, best_cost: 0.04981798346803804, cost_count: 45, time: 0.000148464, action: ContractionInside
Dec 13 16:05:48.476 INFO iter: 24, cost: 0.04981798346803804, best_cost: 0.04981798346803804, cost_count: 47, time: 0.000187549, action: ContractionInside
Dec 13 16:05:48.476 INFO iter: 25, cost: 0.0476407466669301, best_cost: 0.0476407466669301, cost_count: 49, action: ContractionOutside, time: 0.000143021
Dec 13 16:05:48.477 INFO iter: 26, cost: 0.0476407466669301, best_cost: 0.0476407466669301, cost_count: 51, time: 0.000156156, action: ContractionInside
Dec 13 16:05:48.477 INFO iter: 27, cost: 0.0476407466669301, best_cost: 0.0476407466669301, cost_count: 53, time: 0.000177552, action: ContractionOutside
Dec 13 16:05:48.477 INFO iter: 28, cost: 0.047425336008215926, best_cost: 0.047425336008215926, cost_count: 55, action: ContractionInside, time: 0.000167305
Dec 13 16:05:48.478 INFO iter: 29, cost: 0.04613973773410756, best_cost: 0.04613973773410756, cost_count: 57, time: 0.000208689, action: ContractionInside
Dec 13 16:05:48.478 INFO iter: 30, cost: 0.04343181081210659, best_cost: 0.04343181081210659, cost_count: 59, time: 0.000239173, action: Expansion
Dec 13 16:05:48.478 INFO iter: 31, cost: 0.04343181081210659, best_cost: 0.04343181081210659, cost_count: 60, action: Reflection, time: 0.000123339
Dec 13 16:05:48.478 INFO iter: 32, cost: 0.04343181081210659, best_cost: 0.04343181081210659, cost_count: 61, action: Reflection, time: 0.000101168
Dec 13 16:05:48.478 INFO iter: 33, cost: 0.04343181081210659, best_cost: 0.04343181081210659, cost_count: 63, time: 0.00019135, action: ContractionInside
Dec 13 16:05:48.478 INFO iter: 34, cost: 0.04343181081210659, best_cost: 0.04343181081210659, cost_count: 65, action: ContractionInside, time: 0.000147258
Dec 13 16:05:48.478 INFO iter: 35, cost: 0.040365389427234305, best_cost: 0.040365389427234305, cost_count: 67, time: 0.000137332, action: Expansion
Dec 13 16:05:48.478 INFO iter: 36, cost: 0.040365389427234305, best_cost: 0.040365389427234305, cost_count: 68, time: 0.000068571, action: Reflection
Dec 13 16:05:48.478 INFO iter: 37, cost: 0.03592380763438113, best_cost: 0.03592380763438113, cost_count: 70, time: 0.000138786, action: Expansion
Dec 13 16:05:48.479 INFO iter: 38, cost: 0.03592380763438113, best_cost: 0.03592380763438113, cost_count: 71, action: Reflection, time: 0.000087557
Dec 13 16:05:48.479 INFO iter: 39, cost: 0.03592380763438113, best_cost: 0.03592380763438113, cost_count: 73, action: ContractionInside, time: 0.000129377
Dec 13 16:05:48.479 INFO iter: 40, cost: 0.03592380763438113, best_cost: 0.03592380763438113, cost_count: 75, time: 0.000136022, action: ContractionInside
Dec 13 16:05:48.479 INFO iter: 41, cost: 0.03414289694340923, best_cost: 0.03414289694340923, cost_count: 77, time: 0.00012623, action: Expansion
Dec 13 16:05:48.479 INFO iter: 42, cost: 0.03414289694340923, best_cost: 0.03414289694340923, cost_count: 78, time: 0.000085775, action: Reflection
Dec 13 16:05:48.479 INFO iter: 43, cost: 0.03414289694340923, best_cost: 0.03414289694340923, cost_count: 79, action: Reflection, time: 0.0000601
Dec 13 16:05:48.479 INFO iter: 44, cost: 0.03377750883993412, best_cost: 0.03377750883993412, cost_count: 81, time: 0.000115157, action: Expansion
Dec 13 16:05:48.479 INFO iter: 45, cost: 0.03377750883993412, best_cost: 0.03377750883993412, cost_count: 82, action: Reflection, time: 0.000059951
Dec 13 16:05:48.479 INFO iter: 46, cost: 0.03377750883993412, best_cost: 0.03377750883993412, cost_count: 84, time: 0.000119815, action: ContractionInside
Dec 13 16:05:48.480 INFO iter: 47, cost: 0.03377750883993412, best_cost: 0.03377750883993412, cost_count: 86, action: ContractionInside, time: 0.000185798
Dec 13 16:05:48.480 INFO iter: 48, cost: 0.03348361144354422, best_cost: 0.03348361144354422, cost_count: 88, action: Expansion, time: 0.000201059
Dec 13 16:05:48.480 INFO iter: 49, cost: 0.03345080848817214, best_cost: 0.03345080848817214, cost_count: 90, time: 0.000189569, action: Expansion
Dec 13 16:05:48.480 INFO iter: 50, cost: 0.03345080848817214, best_cost: 0.03345080848817214, cost_count: 92, action: ContractionInside, time: 0.000116172
Dec 13 16:05:48.480 INFO iter: 51, cost: 0.033089873861584254, best_cost: 0.033089873861584254, cost_count: 94, action: Expansion, time: 0.000113781
Dec 13 16:05:48.480 INFO iter: 52, cost: 0.03274066065923339, best_cost: 0.03274066065923339, cost_count: 96, action: ContractionOutside, time: 0.0001136
Dec 13 16:05:48.481 INFO iter: 53, cost: 0.03268898427002615, best_cost: 0.03268898427002615, cost_count: 98, time: 0.000114027, action: ContractionInside
Dec 13 16:05:48.481 INFO iter: 54, cost: 0.03268898427002615, best_cost: 0.03268898427002615, cost_count: 100, action: ContractionInside, time: 0.000114045
Dec 13 16:05:48.481 INFO iter: 55, cost: 0.03234418817662415, best_cost: 0.03234418817662415, cost_count: 102, time: 0.000113672, action: ContractionInside
Dec 13 16:05:48.481 INFO iter: 56, cost: 0.03234418817662415, best_cost: 0.03234418817662415, cost_count: 104, time: 0.000114328, action: ContractionInside
Dec 13 16:05:48.481 INFO iter: 57, cost: 0.03234418817662415, best_cost: 0.03234418817662415, cost_count: 105, action: Reflection, time: 0.000059241
Dec 13 16:05:48.481 INFO iter: 58, cost: 0.03234418817662415, best_cost: 0.03234418817662415, cost_count: 106, action: Reflection, time: 0.000059837
Dec 13 16:05:48.481 INFO iter: 59, cost: 0.032276865320111564, best_cost: 0.032276865320111564, cost_count: 108, time: 0.000115052, action: ContractionInside
Dec 13 16:05:48.481 INFO iter: 60, cost: 0.032276865320111564, best_cost: 0.032276865320111564, cost_count: 110, time: 0.000113418, action: ContractionInside
Dec 13 16:05:48.481 INFO iter: 61, cost: 0.03225761497820422, best_cost: 0.03225761497820422, cost_count: 112, time: 0.000113211, action: ContractionOutside
Dec 13 16:05:48.482 INFO iter: 62, cost: 0.03221158705926197, best_cost: 0.03221158705926197, cost_count: 114, action: ContractionInside, time: 0.000113502
Dec 13 16:05:48.482 INFO iter: 63, cost: 0.03221158705926197, best_cost: 0.03221158705926197, cost_count: 115, action: Reflection, time: 0.000059632
Dec 13 16:05:48.482 INFO iter: 64, cost: 0.032156234520969586, best_cost: 0.032156234520969586, cost_count: 117, action: ContractionInside, time: 0.000114247
Dec 13 16:05:48.482 INFO iter: 65, cost: 0.032005385796857806, best_cost: 0.032005385796857806, cost_count: 119, action: Expansion, time: 0.000152021
Dec 13 16:05:48.491 INFO iter: 66, cost: 0.032005385796857806, best_cost: 0.032005385796857806, cost_count: 121, time: 0.000174062, action: ContractionOutside
Dec 13 16:05:48.491 INFO iter: 67, cost: 0.032005385796857806, best_cost: 0.032005385796857806, cost_count: 122, time: 0.000095715, action: Reflection
Dec 13 16:05:48.491 INFO iter: 68, cost: 0.032005385796857806, best_cost: 0.032005385796857806, cost_count: 123, time: 0.00009065, action: Reflection
Dec 13 16:05:48.491 INFO iter: 69, cost: 0.032005385796857806, best_cost: 0.032005385796857806, cost_count: 124, time: 0.000085897, action: Reflection
Dec 13 16:05:48.491 INFO iter: 70, cost: 0.032005385796857806, best_cost: 0.032005385796857806, cost_count: 125, action: Reflection, time: 0.000060648
Dec 13 16:05:48.491 INFO iter: 71, cost: 0.031879006993859045, best_cost: 0.031879006993859045, cost_count: 127, time: 0.000120866, action: Expansion
Dec 13 16:05:48.491 INFO iter: 72, cost: 0.031879006993859045, best_cost: 0.031879006993859045, cost_count: 128, time: 0.000059861, action: Reflection
Dec 13 16:05:48.491 INFO iter: 73, cost: 0.03176166106233146, best_cost: 0.03176166106233146, cost_count: 130, time: 0.000112931, action: Expansion
Dec 13 16:05:48.491 INFO iter: 74, cost: 0.03171393156205579, best_cost: 0.03171393156205579, cost_count: 132, time: 0.000112956, action: Expansion
Dec 13 16:05:48.491 INFO iter: 75, cost: 0.03171393156205579, best_cost: 0.03171393156205579, cost_count: 133, time: 0.000059929, action: Reflection
Dec 13 16:05:48.491 INFO iter: 76, cost: 0.03171393156205579, best_cost: 0.03171393156205579, cost_count: 134, action: Reflection, time: 0.000060233
Dec 13 16:05:48.492 INFO iter: 77, cost: 0.03171393156205579, best_cost: 0.03171393156205579, cost_count: 136, action: ContractionInside, time: 0.00011338
Dec 13 16:05:48.492 INFO iter: 78, cost: 0.03171393156205579, best_cost: 0.03171393156205579, cost_count: 137, time: 0.000058811, action: Reflection
Dec 13 16:05:48.492 INFO iter: 79, cost: 0.031550045087630506, best_cost: 0.031550045087630506, cost_count: 139, time: 0.000114382, action: Expansion
Dec 13 16:05:48.492 INFO iter: 80, cost: 0.031550045087630506, best_cost: 0.031550045087630506, cost_count: 141, time: 0.000118709, action: ContractionOutside
Dec 13 16:05:48.492 INFO iter: 81, cost: 0.031550045087630506, best_cost: 0.031550045087630506, cost_count: 142, action: Reflection, time: 0.000059266
Dec 13 16:05:48.492 INFO iter: 82, cost: 0.031550045087630506, best_cost: 0.031550045087630506, cost_count: 144, time: 0.000112591, action: ContractionInside
Dec 13 16:05:48.492 INFO iter: 83, cost: 0.031550045087630506, best_cost: 0.031550045087630506, cost_count: 146, time: 0.00011309, action: ContractionInside
Dec 13 16:05:48.492 INFO iter: 84, cost: 0.0315040934384468, best_cost: 0.0315040934384468, cost_count: 148, time: 0.000111893, action: Expansion
Dec 13 16:05:48.492 INFO iter: 85, cost: 0.03149400477429251, best_cost: 0.03149400477429251, cost_count: 150, action: Expansion, time: 0.000114632
Dec 13 16:05:48.492 INFO iter: 86, cost: 0.03149400477429251, best_cost: 0.03149400477429251, cost_count: 151, action: Reflection, time: 0.000059106
Dec 13 16:05:48.492 INFO iter: 87, cost: 0.03149400477429251, best_cost: 0.03149400477429251, cost_count: 152, time: 0.000059585, action: Reflection
Dec 13 16:05:48.493 INFO iter: 88, cost: 0.03147959338975672, best_cost: 0.03147959338975672, cost_count: 154, action: Expansion, time: 0.000112926
Dec 13 16:05:48.493 INFO iter: 89, cost: 0.03145269972159686, best_cost: 0.03145269972159686, cost_count: 156, action: Expansion, time: 0.000112655
Dec 13 16:05:48.493 INFO iter: 90, cost: 0.03145269972159686, best_cost: 0.03145269972159686, cost_count: 158, time: 0.000112987, action: ContractionInside
Dec 13 16:05:48.493 INFO iter: 91, cost: 0.03144054134572604, best_cost: 0.03144054134572604, cost_count: 160, time: 0.000112514, action: Expansion
Dec 13 16:05:48.493 INFO iter: 92, cost: 0.03144054134572604, best_cost: 0.03144054134572604, cost_count: 162, action: ContractionInside, time: 0.000113071
Dec 13 16:05:48.493 INFO iter: 93, cost: 0.03137949677661753, best_cost: 0.03137949677661753, cost_count: 164, time: 0.000118989, action: Expansion
Dec 13 16:05:48.493 INFO iter: 94, cost: 0.03137949677661753, best_cost: 0.03137949677661753, cost_count: 166, time: 0.000113693, action: ContractionInside
Dec 13 16:05:48.493 INFO iter: 95, cost: 0.03137949677661753, best_cost: 0.03137949677661753, cost_count: 167, action: Reflection, time: 0.00005862
Dec 13 16:05:48.493 INFO iter: 96, cost: 0.03137949677661753, best_cost: 0.03137949677661753, cost_count: 168, action: Reflection, time: 0.000058745
Dec 13 16:05:48.493 INFO iter: 97, cost: 0.03137949677661753, best_cost: 0.03137949677661753, cost_count: 169, action: Reflection, time: 0.000059016
Dec 13 16:05:48.493 INFO iter: 98, cost: 0.03137949677661753, best_cost: 0.03137949677661753, cost_count: 170, action: Reflection, time: 0.000058706
Dec 13 16:05:48.493 INFO iter: 99, cost: 0.03137949677661753, best_cost: 0.03137949677661753, cost_count: 172, action: ContractionInside, time: 0.000113346
Dec 13 16:05:48.493 INFO iter: 100, cost: 0.03137949677661753, best_cost: 0.03137949677661753, cost_count: 174, time: 0.000111745, action: ContractionOutside
Dec 13 16:05:48.494 INFO iter: 101, cost: 0.03137949677661753, best_cost: 0.03137949677661753, cost_count: 175, time: 0.000058709, action: Reflection
Dec 13 16:05:48.494 INFO iter: 102, cost: 0.03137949677661753, best_cost: 0.03137949677661753, cost_count: 176, time: 0.000058953, action: Reflection
Dec 13 16:05:48.494 INFO iter: 103, cost: 0.03132379645707613, best_cost: 0.03132379645707613, cost_count: 178, action: Expansion, time: 0.000111973
Dec 13 16:05:48.494 INFO iter: 104, cost: 0.03132379645707613, best_cost: 0.03132379645707613, cost_count: 179, time: 0.000058558, action: Reflection
Dec 13 16:05:48.494 INFO iter: 105, cost: 0.03132379645707613, best_cost: 0.03132379645707613, cost_count: 180, action: Reflection, time: 0.00005816
Dec 13 16:05:48.494 INFO iter: 106, cost: 0.03132379645707613, best_cost: 0.03132379645707613, cost_count: 182, action: ContractionInside, time: 0.000112088
Dec 13 16:05:48.494 INFO iter: 107, cost: 0.03132379645707613, best_cost: 0.03132379645707613, cost_count: 183, action: Reflection, time: 0.000058276
Dec 13 16:05:48.494 INFO iter: 108, cost: 0.031297913313495164, best_cost: 0.031297913313495164, cost_count: 185, action: Expansion, time: 0.000112287
Dec 13 16:05:48.494 INFO iter: 109, cost: 0.031297913313495164, best_cost: 0.031297913313495164, cost_count: 186, time: 0.000058827, action: Reflection
Dec 13 16:05:48.494 INFO iter: 110, cost: 0.031297913313495164, best_cost: 0.031297913313495164, cost_count: 188, action: ContractionInside, time: 0.0001118
Dec 13 16:05:48.494 INFO iter: 111, cost: 0.031297913313495164, best_cost: 0.031297913313495164, cost_count: 189, action: Reflection, time: 0.000058389
Dec 13 16:05:48.494 INFO iter: 112, cost: 0.031221833802620853, best_cost: 0.031221833802620853, cost_count: 191, time: 0.000112385, action: Expansion
Dec 13 16:05:48.494 INFO iter: 113, cost: 0.031221833802620853, best_cost: 0.031221833802620853, cost_count: 192, action: Reflection, time: 0.000058605
Dec 13 16:05:48.495 INFO iter: 114, cost: 0.031221833802620853, best_cost: 0.031221833802620853, cost_count: 193, action: Reflection, time: 0.000058512
Dec 13 16:05:48.495 INFO iter: 115, cost: 0.031221833802620853, best_cost: 0.031221833802620853, cost_count: 194, time: 0.000058747, action: Reflection
Dec 13 16:05:48.495 INFO iter: 116, cost: 0.03116654344292932, best_cost: 0.03116654344292932, cost_count: 196, action: Expansion, time: 0.000121207
Dec 13 16:05:48.495 INFO iter: 117, cost: 0.03116654344292932, best_cost: 0.03116654344292932, cost_count: 197, time: 0.000058884, action: Reflection
Dec 13 16:05:48.495 INFO iter: 118, cost: 0.03116338971440618, best_cost: 0.03116338971440618, cost_count: 199, time: 0.000111846, action: Expansion
Dec 13 16:05:48.495 INFO iter: 119, cost: 0.03110737733686549, best_cost: 0.03110737733686549, cost_count: 201, time: 0.000112369, action: Expansion
Dec 13 16:05:48.495 INFO iter: 120, cost: 0.03109075066447514, best_cost: 0.03109075066447514, cost_count: 203, time: 0.000112726, action: Expansion
Dec 13 16:05:48.495 INFO iter: 121, cost: 0.031039400782249243, best_cost: 0.031039400782249243, cost_count: 205, action: Expansion, time: 0.000113236
Dec 13 16:05:48.495 INFO iter: 122, cost: 0.031039400782249243, best_cost: 0.031039400782249243, cost_count: 206, time: 0.000058569, action: Reflection
Dec 13 16:05:48.495 INFO iter: 123, cost: 0.03101420881979535, best_cost: 0.03101420881979535, cost_count: 208, time: 0.000111923, action: Expansion
Dec 13 16:05:48.495 INFO iter: 124, cost: 0.030951234457747407, best_cost: 0.030951234457747407, cost_count: 210, action: Expansion, time: 0.000111966
Dec 13 16:05:48.495 INFO iter: 125, cost: 0.030951234457747407, best_cost: 0.030951234457747407, cost_count: 211, time: 0.000059148, action: Reflection
Dec 13 16:05:48.495 INFO iter: 126, cost: 0.030951234457747407, best_cost: 0.030951234457747407, cost_count: 213, action: ContractionInside, time: 0.00011254
Dec 13 16:05:48.496 INFO iter: 127, cost: 0.030924713248677148, best_cost: 0.030924713248677148, cost_count: 215, action: Expansion, time: 0.000112993
Dec 13 16:05:48.496 INFO iter: 128, cost: 0.030924713248677148, best_cost: 0.030924713248677148, cost_count: 217, time: 0.000112471, action: ContractionOutside
Dec 13 16:05:48.496 INFO iter: 129, cost: 0.030924713248677148, best_cost: 0.030924713248677148, cost_count: 218, time: 0.000058947, action: Reflection
Dec 13 16:05:48.496 INFO iter: 130, cost: 0.030924713248677148, best_cost: 0.030924713248677148, cost_count: 220, action: ContractionOutside, time: 0.000112457
Dec 13 16:05:48.496 INFO iter: 131, cost: 0.030924713248677148, best_cost: 0.030924713248677148, cost_count: 222, action: ContractionOutside, time: 0.000112544
Dec 13 16:05:48.496 INFO iter: 132, cost: 0.03091403723604345, best_cost: 0.03091403723604345, cost_count: 224, action: Expansion, time: 0.000113116
Dec 13 16:05:48.496 INFO iter: 133, cost: 0.03091403723604345, best_cost: 0.03091403723604345, cost_count: 226, time: 0.000112752, action: ContractionInside
Dec 13 16:05:48.496 INFO iter: 134, cost: 0.030910203110480046, best_cost: 0.030910203110480046, cost_count: 228, action: Expansion, time: 0.000112736
Dec 13 16:05:48.496 INFO iter: 135, cost: 0.030905209770135316, best_cost: 0.030905209770135316, cost_count: 230, time: 0.000113438, action: Expansion
Dec 13 16:05:48.496 INFO iter: 136, cost: 0.03090184749772371, best_cost: 0.03090184749772371, cost_count: 232, action: Expansion, time: 0.000113515
Dec 13 16:05:48.496 INFO iter: 137, cost: 0.03090184749772371, best_cost: 0.03090184749772371, cost_count: 233, action: Reflection, time: 0.000059
Dec 13 16:05:48.496 INFO iter: 138, cost: 0.030888401850203414, best_cost: 0.030888401850203414, cost_count: 235, time: 0.000111041, action: Expansion
Dec 13 16:05:48.496 INFO iter: 139, cost: 0.030888401850203414, best_cost: 0.030888401850203414, cost_count: 236, time: 0.00005762, action: Reflection
Dec 13 16:05:48.497 INFO iter: 140, cost: 0.030876446582383105, best_cost: 0.030876446582383105, cost_count: 238, action: Expansion, time: 0.000111968
Dec 13 16:05:48.497 INFO iter: 141, cost: 0.030876446582383105, best_cost: 0.030876446582383105, cost_count: 239, time: 0.000058987, action: Reflection
Dec 13 16:05:48.497 INFO iter: 142, cost: 0.030876446582383105, best_cost: 0.030876446582383105, cost_count: 240, time: 0.000058844, action: Reflection
Dec 13 16:05:48.497 INFO iter: 143, cost: 0.030876446582383105, best_cost: 0.030876446582383105, cost_count: 241, action: Reflection, time: 0.000058508
Dec 13 16:05:48.497 INFO iter: 144, cost: 0.030876446582383105, best_cost: 0.030876446582383105, cost_count: 242, time: 0.000058478, action: Reflection
Dec 13 16:05:48.497 INFO iter: 145, cost: 0.030876446582383105, best_cost: 0.030876446582383105, cost_count: 243, time: 0.000058734, action: Reflection
Dec 13 16:05:48.497 INFO iter: 146, cost: 0.030867766108365974, best_cost: 0.030867766108365974, cost_count: 245, time: 0.000112446, action: Expansion
Dec 13 16:05:48.497 INFO iter: 147, cost: 0.030867766108365974, best_cost: 0.030867766108365974, cost_count: 246, action: Reflection, time: 0.000059771
Dec 13 16:05:48.497 INFO iter: 148, cost: 0.030867766108365974, best_cost: 0.030867766108365974, cost_count: 248, time: 0.000112674, action: ContractionInside
Dec 13 16:05:48.497 INFO iter: 149, cost: 0.030867766108365974, best_cost: 0.030867766108365974, cost_count: 249, action: Reflection, time: 0.000059334
Dec 13 16:05:48.511 INFO iter: 150, cost: 0.030867766108365974, best_cost: 0.030867766108365974, cost_count: 251, time: 0.000123883, action: ContractionInside
Dec 13 16:05:48.511 INFO iter: 151, cost: 0.030867604498606586, best_cost: 0.030867604498606586, cost_count: 253, time: 0.000112339, action: ContractionInside
Dec 13 16:05:48.511 INFO iter: 152, cost: 0.030867604498606586, best_cost: 0.030867604498606586, cost_count: 254, time: 0.00005864, action: Reflection
Dec 13 16:05:48.511 INFO iter: 153, cost: 0.030867604498606586, best_cost: 0.030867604498606586, cost_count: 255, time: 0.00005897, action: Reflection
Dec 13 16:05:48.511 INFO iter: 154, cost: 0.030866373551625092, best_cost: 0.030866373551625092, cost_count: 257, time: 0.000113394, action: ContractionInside
Dec 13 16:05:48.511 INFO iter: 155, cost: 0.03086603392042653, best_cost: 0.03086603392042653, cost_count: 259, action: ContractionInside, time: 0.000112343
Dec 13 16:05:48.512 INFO iter: 156, cost: 0.030865494381929842, best_cost: 0.030865494381929842, cost_count: 261, action: ContractionInside, time: 0.000136331
Dec 13 16:05:48.512 INFO iter: 157, cost: 0.030865161332494874, best_cost: 0.030865161332494874, cost_count: 263, time: 0.000151266, action: ContractionOutside
Dec 13 16:05:48.512 INFO iter: 158, cost: 0.03086486288500786, best_cost: 0.03086486288500786, cost_count: 265, time: 0.000122158, action: ContractionInside
thread 'main' panicked at /root/.cargo/registry/src/index.crates.io-6f17d22bba15001f/argmin-0.8.1/src/solver/neldermead/mod.rs:355:43:
attempt to subtract with overflow
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace

[stefan-k]

Thanks. This is a strange behaviour. Does it always happen at the same number of iterations? Could you run it with RUST_BACKTRACE=1 to see the backtrace, please?

[luxiant]

Oh, I was an idiot. You are right that this code is not wrong. It is another nelder mead calculation that causes an error because of the parameter vector of which the number is less than required. I didn't notice that part.
Anyway, thanks! I finally finished my project.

[stefan-k]

:) No worries, happy to see you got it fixed!