Shuffled rows

match ← All logistic growth benchmarks

The basic history in a scrambled order. Both libraries must sort it the same way.

Input

benchmark/inputs/v2/logistic-unsorted-96.json: 96 history rows and 120 forecast rows.

Output

Largest difference: 2.4e-11 (< 0.0001% of the input's range).

Every forecast row from both libraries.

Date	Effect Prophet	Python Prophet	Difference
2020-01-01	14.1879	14.1879	2.2e-13
2020-01-02	14.7199	14.7199	8.9e-15
2020-01-03	15.2686	15.2686	2.1e-13
2020-01-04	15.8340	15.8340	4.6e-13
2020-01-05	16.4165	16.4165	7.1e-13
2020-01-06	17.0162	17.0162	9.7e-13
2020-01-07	17.6335	17.6335	1.3e-12
2020-01-08	18.2684	18.2684	1.6e-12
2020-01-09	18.9213	18.9213	1.9e-12
2020-01-10	19.5921	19.5921	2.2e-12
2020-01-11	20.2811	20.2811	2.5e-12
2020-01-12	20.9883	20.9883	2.9e-12
2020-01-13	21.7138	21.7138	3.3e-12
2020-01-14	22.4577	22.4577	3.7e-12
2020-01-15	23.2199	23.2199	4.1e-12
2020-01-16	24.0005	24.0005	4.5e-12
2020-01-17	24.7993	24.7993	4.9e-12
2020-01-18	25.6163	25.6163	5.3e-12
2020-01-19	26.4514	26.4514	5.8e-12
2020-01-20	27.3043	27.3043	6.3e-12
2020-01-21	28.1749	28.1749	6.8e-12
2020-01-22	29.0629	29.0629	7.2e-12
2020-01-23	29.9681	29.9681	7.7e-12
2020-01-24	30.8900	30.8900	8.3e-12
2020-01-25	31.8284	31.8284	8.8e-12
2020-01-26	32.7827	32.7827	9.3e-12
2020-01-27	33.7526	33.7526	9.9e-12
2020-01-28	34.7376	34.7376	1.0e-11
2020-01-29	35.7371	35.7371	1.1e-11
2020-01-30	36.7504	36.7504	1.2e-11
2020-01-31	37.7771	37.7771	1.2e-11
2020-02-01	38.8165	38.8165	1.3e-11
2020-02-02	39.8677	39.8677	1.3e-11
2020-02-03	40.9302	40.9302	1.4e-11
2020-02-04	42.0031	42.0031	1.4e-11
2020-02-05	43.0857	43.0857	1.5e-11
2020-02-06	44.1771	44.1771	1.5e-11
2020-02-07	45.2764	45.2764	1.6e-11
2020-02-08	46.3828	46.3828	1.6e-11
2020-02-09	47.4958	47.4958	1.7e-11
2020-02-10	48.6140	48.6140	1.8e-11
2020-02-11	49.7365	49.7365	1.8e-11
2020-02-12	50.8624	50.8624	1.9e-11
2020-02-13	51.9907	51.9907	1.9e-11
2020-02-14	53.1204	53.1204	1.9e-11
2020-02-15	54.2507	54.2507	2.0e-11
2020-02-16	55.3804	55.3804	2.0e-11
2020-02-17	56.5087	56.5087	2.1e-11
2020-02-18	57.6346	57.6346	2.1e-11
2020-02-19	58.7571	58.7571	2.1e-11
2020-02-20	59.8754	59.8754	2.2e-11
2020-02-21	60.9885	60.9885	2.2e-11
2020-02-22	62.0955	62.0955	2.2e-11
2020-02-23	63.1956	63.1956	2.3e-11
2020-02-24	64.2880	64.2880	2.3e-11
2020-02-25	65.3718	65.3718	2.3e-11
2020-02-26	66.4462	66.4462	2.3e-11
2020-02-27	67.5105	67.5105	2.4e-11
2020-02-28	68.5640	68.5640	2.4e-11
2020-02-29	69.6060	69.6060	2.4e-11
2020-03-01	70.6359	70.6359	2.4e-11
2020-03-02	71.6530	71.6530	2.4e-11
2020-03-03	72.6569	72.6569	2.4e-11
2020-03-04	73.6470	73.6470	2.4e-11
2020-03-05	74.6227	74.6227	2.4e-11
2020-03-06	75.5838	75.5838	2.4e-11
2020-03-07	76.5297	76.5297	2.4e-11
2020-03-08	77.4601	77.4601	2.4e-11
2020-03-09	78.3748	78.3748	2.4e-11
2020-03-10	79.2734	79.2734	2.4e-11
2020-03-11	80.1556	80.1556	2.3e-11
2020-03-12	81.0214	81.0214	2.3e-11
2020-03-13	81.8705	81.8705	2.3e-11
2020-03-14	82.7028	82.7028	2.3e-11
2020-03-15	83.5182	83.5182	2.3e-11
2020-03-16	84.3166	84.3166	2.2e-11
2020-03-17	85.0981	85.0981	2.2e-11
2020-03-18	85.8626	85.8626	2.2e-11
2020-03-19	86.6102	86.6102	2.2e-11
2020-03-20	87.3409	87.3409	2.1e-11
2020-03-21	88.0548	88.0548	2.1e-11
2020-03-22	88.7520	88.7520	2.1e-11
2020-03-23	89.4327	89.4327	2.1e-11
2020-03-24	90.0970	90.0970	2.0e-11
2020-03-25	90.7451	90.7451	2.0e-11
2020-03-26	91.3772	91.3772	2.0e-11
2020-03-27	91.9934	91.9934	1.9e-11
2020-03-28	92.5941	92.5941	1.9e-11
2020-03-29	93.1795	93.1795	1.9e-11
2020-03-30	93.7497	93.7497	1.8e-11
2020-03-31	94.3051	94.3051	1.8e-11
2020-04-01	94.8460	94.8460	1.7e-11
2020-04-02	95.3725	95.3725	1.7e-11
2020-04-03	95.8851	95.8851	1.7e-11
2020-04-04	96.3839	96.3839	1.6e-11
2020-04-05	96.8694	96.8694	1.6e-11
2020-04-06	97.3417	97.3417	1.6e-11
2020-04-07	97.8012	97.8012	1.5e-11
2020-04-08	98.2483	98.2483	1.5e-11
2020-04-09	98.6831	98.6831	1.5e-11
2020-04-10	99.1060	99.1060	1.4e-11
2020-04-11	99.5174	99.5174	1.4e-11
2020-04-12	99.9175	99.9175	1.4e-11
2020-04-13	100.3066	100.3066	1.3e-11
2020-04-14	100.6851	100.6851	1.3e-11
2020-04-15	101.0532	101.0532	1.3e-11
2020-04-16	101.4112	101.4112	1.2e-11
2020-04-17	101.7594	101.7594	1.2e-11
2020-04-18	102.0981	102.0981	1.2e-11
2020-04-19	102.4276	102.4276	1.1e-11
2020-04-20	102.7482	102.7482	1.1e-11
2020-04-21	103.0602	103.0602	1.1e-11
2020-04-22	103.3637	103.3637	1.0e-11
2020-04-23	103.6592	103.6592	1.0e-11
2020-04-24	103.9468	103.9468	9.8e-12
2020-04-25	104.2267	104.2267	9.6e-12
2020-04-26	104.4993	104.4993	9.3e-12
2020-04-27	104.7648	104.7648	9.0e-12
2020-04-28	105.0234	105.0234	8.7e-12
2020-04-29	105.2754	105.2754	8.5e-12

Timing

Median of all measured repetitions.

Step	Effect Prophet	Python Prophet	Effect Prophet is
Fit	8.28 ms	17.9 ms	2.2× faster
Predict	842 µs	3.22 ms	3.8× faster
Save model to JSON	19 µs	435 µs	22× faster
Load model from JSON	134 µs	2.49 ms	19× faster
Load and predict in a new process	158 ms	1.62 s	10× faster

Settings

The case declaration both libraries ran.

{
  "kind": "map",
  "comparison": {
    "kind": "equivalent-objective",
    "evidenceId": "logistic-python-reconciliation-v1"
  },
  "configuration": {
    "growth": "logistic",
    "scaling": "absmax",
    "seasonalityMode": "additive",
    "holidaysMode": "additive",
    "changepoints": {
      "mode": "explicit",
      "timestamps": [
        "2020-02-08T00:00:00.000Z"
      ]
    },
    "changepointPriorScale": 0.05,
    "seasonalities": [],
    "events": [],
    "regressors": []
  },
  "pythonOptimizer": {
    "algorithm": "Auto",
    "maxIterations": 10000,
    "newtonFallback": true,
    "sigFigs": 12
  }
}