No changepoints, minmax scaling

match ← All logistic growth benchmarks

96 daily rows rising toward a ceiling of about 100, with a little noise. Prediction covers the history and 24 more days. The curve's rate never changes. Values are scaled between the lowest floor and the highest ceiling.

Input

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

Output

Largest difference: 1.6e-10 (< 0.0001% of the input's range).

Every forecast row from both libraries.

Date	Effect Prophet	Python Prophet	Difference
2020-01-01	20.0599	20.0599	2.3e-11
2020-01-02	20.3521	20.3521	2.4e-11
2020-01-03	20.6579	20.6579	2.5e-11
2020-01-04	20.9776	20.9776	2.7e-11
2020-01-05	21.3119	21.3119	2.9e-11
2020-01-06	21.6612	21.6612	3.0e-11
2020-01-07	22.0261	22.0261	3.2e-11
2020-01-08	22.4070	22.4070	3.4e-11
2020-01-09	22.8046	22.8046	3.5e-11
2020-01-10	23.2194	23.2194	3.7e-11
2020-01-11	23.6519	23.6519	3.9e-11
2020-01-12	24.1025	24.1025	4.1e-11
2020-01-13	24.5720	24.5720	4.4e-11
2020-01-14	25.0607	25.0607	4.6e-11
2020-01-15	25.5692	25.5692	4.8e-11
2020-01-16	26.0979	26.0979	5.1e-11
2020-01-17	26.6474	26.6474	5.3e-11
2020-01-18	27.2181	27.2181	5.6e-11
2020-01-19	27.8104	27.8104	5.8e-11
2020-01-20	28.4247	28.4247	6.1e-11
2020-01-21	29.0614	29.0614	6.4e-11
2020-01-22	29.7209	29.7209	6.7e-11
2020-01-23	30.4033	30.4033	7.0e-11
2020-01-24	31.1091	31.1091	7.3e-11
2020-01-25	31.8384	31.8384	7.6e-11
2020-01-26	32.5913	32.5913	7.9e-11
2020-01-27	33.3679	33.3679	8.2e-11
2020-01-28	34.1684	34.1684	8.6e-11
2020-01-29	34.9926	34.9926	8.9e-11
2020-01-30	35.8405	35.8405	9.2e-11
2020-01-31	36.7120	36.7120	9.6e-11
2020-02-01	37.6068	37.6068	9.9e-11
2020-02-02	38.5247	38.5247	1.0e-10
2020-02-03	39.4651	39.4651	1.1e-10
2020-02-04	40.4278	40.4278	1.1e-10
2020-02-05	41.4121	41.4121	1.1e-10
2020-02-06	42.4175	42.4175	1.2e-10
2020-02-07	43.4432	43.4432	1.2e-10
2020-02-08	44.4884	44.4884	1.2e-10
2020-02-09	45.5524	45.5524	1.3e-10
2020-02-10	46.6341	46.6341	1.3e-10
2020-02-11	47.7325	47.7325	1.3e-10
2020-02-12	48.8466	48.8466	1.3e-10
2020-02-13	49.9751	49.9751	1.4e-10
2020-02-14	51.1169	51.1169	1.4e-10
2020-02-15	52.2707	52.2707	1.4e-10
2020-02-16	53.4351	53.4351	1.4e-10
2020-02-17	54.6087	54.6087	1.5e-10
2020-02-18	55.7901	55.7901	1.5e-10
2020-02-19	56.9778	56.9778	1.5e-10
2020-02-20	58.1703	58.1703	1.5e-10
2020-02-21	59.3662	59.3662	1.5e-10
2020-02-22	60.5638	60.5638	1.5e-10
2020-02-23	61.7616	61.7616	1.5e-10
2020-02-24	62.9582	62.9582	1.5e-10
2020-02-25	64.1520	64.1520	1.6e-10
2020-02-26	65.3414	65.3414	1.6e-10
2020-02-27	66.5251	66.5251	1.6e-10
2020-02-28	67.7015	67.7015	1.6e-10
2020-02-29	68.8694	68.8694	1.6e-10
2020-03-01	70.0273	70.0273	1.5e-10
2020-03-02	71.1740	71.1740	1.5e-10
2020-03-03	72.3081	72.3081	1.5e-10
2020-03-04	73.4287	73.4287	1.5e-10
2020-03-05	74.5345	74.5345	1.5e-10
2020-03-06	75.6246	75.6246	1.5e-10
2020-03-07	76.6980	76.6980	1.5e-10
2020-03-08	77.7538	77.7538	1.5e-10
2020-03-09	78.7912	78.7912	1.4e-10
2020-03-10	79.8096	79.8096	1.4e-10
2020-03-11	80.8082	80.8082	1.4e-10
2020-03-12	81.7866	81.7866	1.4e-10
2020-03-13	82.7441	82.7441	1.4e-10
2020-03-14	83.6806	83.6806	1.3e-10
2020-03-15	84.5956	84.5956	1.3e-10
2020-03-16	85.4888	85.4888	1.3e-10
2020-03-17	86.3601	86.3601	1.2e-10
2020-03-18	87.2093	87.2093	1.2e-10
2020-03-19	88.0365	88.0365	1.2e-10
2020-03-20	88.8416	88.8416	1.2e-10
2020-03-21	89.6246	89.6246	1.1e-10
2020-03-22	90.3858	90.3858	1.1e-10
2020-03-23	91.1252	91.1252	1.1e-10
2020-03-24	91.8430	91.8430	1.0e-10
2020-03-25	92.5396	92.5396	1.0e-10
2020-03-26	93.2151	93.2151	9.9e-11
2020-03-27	93.8700	93.8700	9.6e-11
2020-03-28	94.5045	94.5045	9.3e-11
2020-03-29	95.1191	95.1191	9.0e-11
2020-03-30	95.7142	95.7142	8.7e-11
2020-03-31	96.2901	96.2901	8.4e-11
2020-04-01	96.8473	96.8473	8.2e-11
2020-04-02	97.3863	97.3863	7.9e-11
2020-04-03	97.9075	97.9075	7.6e-11
2020-04-04	98.4114	98.4114	7.4e-11
2020-04-05	98.8985	98.8985	7.1e-11
2020-04-06	99.3693	99.3693	6.9e-11
2020-04-07	99.8243	99.8243	6.6e-11
2020-04-08	100.2639	100.2639	6.4e-11
2020-04-09	100.6886	100.6886	6.2e-11
2020-04-10	101.0990	101.0990	5.9e-11
2020-04-11	101.4954	101.4954	5.7e-11
2020-04-12	101.8785	101.8785	5.5e-11
2020-04-13	102.2485	102.2485	5.3e-11
2020-04-14	102.6061	102.6061	5.1e-11
2020-04-15	102.9517	102.9517	4.9e-11
2020-04-16	103.2857	103.2857	4.7e-11
2020-04-17	103.6085	103.6085	4.5e-11
2020-04-18	103.9206	103.9206	4.3e-11
2020-04-19	104.2224	104.2224	4.2e-11
2020-04-20	104.5143	104.5143	4.0e-11
2020-04-21	104.7967	104.7967	3.8e-11
2020-04-22	105.0699	105.0699	3.7e-11
2020-04-23	105.3344	105.3344	3.5e-11
2020-04-24	105.5905	105.5905	3.4e-11
2020-04-25	105.8385	105.8385	3.3e-11
2020-04-26	106.0789	106.0789	3.1e-11
2020-04-27	106.3118	106.3118	3.0e-11
2020-04-28	106.5377	106.5377	2.9e-11
2020-04-29	106.7569	106.7569	2.7e-11

Timing

Median of all measured repetitions.

Step	Effect Prophet	Python Prophet	Effect Prophet is
Fit	8.51 ms	15.8 ms	1.9× faster
Predict	656 µs	2.98 ms	4.5× faster
Save model to JSON	27 µs	424 µs	16× faster
Load model from JSON	145 µs	2.07 ms	14× faster
Load and predict in a new process	155 ms	1.47 s	9.4× faster

Settings

The case declaration both libraries ran.

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