DailyFcst

import pandas as pd
import numpy as np
import seaborn as sns
from scipy import stats
import matplotlib as mpl
import matplotlib.pyplot as plt

import scipy as stats

from fbprophet import Prophet

The Data

We will be looking at Transactions at advertised locations for the last 2 years, by Day & Week. This data will be used to forecast out weekly Transactions to compare last year to.

data = pd.read_csv('Sales_by_day.csv', parse_dates=['ds'])
txns = pd.DataFrame(data[['ds','# Txns']])
txns['y'] = np.log10(txns['# Txns'])
del txns['# Txns']
txns

data = pd.read_csv('Sales_by_day.csv', parse_dates=['ds'])
sales = pd.DataFrame(data[['ds','Sales $']])
sales['y'] = np.log10(sales['Sales $'])
del sales['Sales $']
sales

	ds	y
0	2015-02-01	4.721918
1	2015-02-02	4.709092
2	2015-02-03	4.673058
3	2015-02-04	4.684917
4	2015-02-05	4.679173
5	2015-02-06	4.755204
6	2015-02-07	4.795623
7	2015-02-08	4.703704
8	2015-02-09	4.667985
9	2015-02-10	4.689362
10	2015-02-11	4.741286
11	2015-02-12	4.801952
12	2015-02-14	4.885039
13	2015-02-15	4.727509
14	2015-02-16	4.664567
15	2015-02-17	4.575142
16	2015-02-18	4.675934
17	2015-02-19	4.705932
18	2015-02-20	4.746580
19	2015-02-21	4.782967
20	2015-02-22	4.680843
21	2015-02-23	4.652672
22	2015-02-24	4.674503
23	2015-02-25	4.713793
24	2015-02-26	4.660078
25	2015-02-27	4.807440
26	2015-02-28	4.831543
27	2015-03-01	4.713742
28	2015-03-02	4.695368
29	2015-03-03	4.692600
...	...	...
695	2017-01-27	4.742937
696	2017-01-28	4.754975
697	2017-01-29	4.647305
698	2017-01-30	4.581665
699	2017-01-31	4.651123
700	2017-02-01	4.753476
701	2017-02-02	4.735224
702	2017-02-03	4.785180
703	2017-02-04	4.825517
704	2017-02-05	4.686538
705	2017-02-06	4.654658
706	2017-02-07	4.729343

	ds	y
0	2015-02-01	6.568501
1	2015-02-02	6.519151
2	2015-02-03	6.352087
3	2015-02-04	6.407623
4	2015-02-05	6.376327
5	2015-02-06	6.469397
6	2015-02-07	6.605302
7	2015-02-08	6.477032
8	2015-02-09	6.347584
9	2015-02-10	6.378210
10	2015-02-11	6.511220
11	2015-02-12	6.560980
12	2015-02-14	6.750798
13	2015-02-15	6.565547
14	2015-02-16	6.469702
15	2015-02-17	6.241425
16	2015-02-18	6.361745
17	2015-02-19	6.426021
18	2015-02-20	6.479443
19	2015-02-21	6.595187
20	2015-02-22	6.457410
21	2015-02-23	6.306829
22	2015-02-24	6.349313
23	2015-02-25	6.431925
24	2015-02-26	6.399252
25	2015-02-27	6.562656
26	2015-02-28	6.667724
27	2015-03-01	6.529803
28	2015-03-02	6.372163
29	2015-03-03	6.389169
...	...	...
695	2017-01-27	6.447965
696	2017-01-28	6.556622
697	2017-01-29	6.405705
698	2017-01-30	6.196282
699	2017-01-31	6.315664
700	2017-02-01	6.568835
701	2017-02-02	6.512520
702	2017-02-03	6.578559
703	2017-02-04	6.701477
704	2017-02-05	6.510305
705	2017-02-06	6.384240
706	2017-02-07	6.501414
707	2017-02-08	6.317689
708	2017-02-09	6.320989
709	2017-02-10	6.443153
710	2017-02-11	6.547922
711	2017-02-12	6.437922
712	2017-02-13	6.472022
713	2017-02-14	6.534376
714	2017-02-15	6.403736
715	2017-02-16	6.382633
716	2017-02-17	6.527329
717	2017-02-18	6.576850
718	2017-02-19	6.470435
719	2017-02-20	6.405650
720	2017-02-21	6.286527
721	2017-02-22	6.500512
722	2017-02-23	6.476577
723	2017-02-24	6.548398
724	2017-02-25	6.612316

725 rows × 2 columns

#forecasting tool used to fit model & plot data
m = Prophet()
m.fit(txns)

<fbprophet.forecaster.Prophet at 0x7f7009d14748>

#create future dataframe with number of periods to be forecasted
future = m.make_future_dataframe(periods=336)
future

	ds
0	2015-02-01
1	2015-02-02
2	2015-02-03
3	2015-02-04
4	2015-02-05
5	2015-02-06
6	2015-02-07
7	2015-02-08
8	2015-02-09
9	2015-02-10
10	2015-02-11
11	2015-02-12
12	2015-02-14
13	2015-02-15
14	2015-02-16
15	2015-02-17
16	2015-02-18
17	2015-02-19
18	2015-02-20
19	2015-02-21
20	2015-02-22
21	2015-02-23
22	2015-02-24
23	2015-02-25
24	2015-02-26
25	2015-02-27
26	2015-02-28
27	2015-03-01
28	2015-03-02
29	2015-03-03
...	...
1031	2017-12-29
1032	2017-12-30
1033	2017-12-31
1034	2018-01-01
1035	2018-01-02
1036	2018-01-03
1037	2018-01-04
1038	2018-01-05
1039	2018-01-06
1040	2018-01-07
1041	2018-01-08
1042	2018-01-09
1043	2018-01-10
1044	2018-01-11
1045	2018-01-12
1046	2018-01-13
1047	2018-01-14
1048	2018-01-15
1049	2018-01-16
1050	2018-01-17
1051	2018-01-18
1052	2018-01-19
1053	2018-01-20
1054	2018-01-21
1055	2018-01-22
1056	2018-01-23
1057	2018-01-24
1058	2018-01-25
1059	2018-01-26
1060	2018-01-27

1061 rows × 1 columns

Initial Forecast

No additional information considered here i.e., Events or Holidays

forecast = m.predict(future)
forecast[['ds', 'yhat', 'yhat_lower', 'yhat_upper']].tail()

	ds	yhat	yhat_lower	yhat_upper
1056	2018-01-23	4.622200	4.560516	4.680854
1057	2018-01-24	4.660621	4.596642	4.721010
1058	2018-01-25	4.662980	4.599900	4.724595
1059	2018-01-26	4.723546	4.664982	4.786545
1060	2018-01-27	4.729829	4.665482	4.793615

%matplotlib inline
m.plot(forecast);

m.plot_components(forecast);

forecast

	ds	t	trend	seasonal_lower	seasonal_upper	trend_lower	trend_upper	yhat_lower	yhat_upper	weekly	weekly_lower	weekly_upper	yearly	yearly_lower	yearly_upper	seasonal	yhat
0	2015-02-01	0.000000	4.750702	-0.060431	-0.060431	4.750702	4.750702	4.629934	4.746744	-0.037397	-0.037397	-0.037397	-0.023033	-0.023033	-0.023033	-0.060431	4.690271
1	2015-02-02	0.001325	4.750573	-0.079103	-0.079103	4.750573	4.750573	4.614538	4.737126	-0.058506	-0.058506	-0.058506	-0.020598	-0.020598	-0.020598	-0.079103	4.671470
2	2015-02-03	0.002649	4.750445	-0.049399	-0.049399	4.750445	4.750445	4.639142	4.758367	-0.030829	-0.030829	-0.030829	-0.018570	-0.018570	-0.018570	-0.049399	4.701046
3	2015-02-04	0.003974	4.750316	-0.013089	-0.013089	4.750316	4.750316	4.679305	4.796501	0.003878	0.003878	0.003878	-0.016966	-0.016966	-0.016966	-0.013089	4.737227
4	2015-02-05	0.005298	4.750187	-0.013413	-0.013413	4.750187	4.750187	4.678724	4.798790	0.002379	0.002379	0.002379	-0.015792	-0.015792	-0.015792	-0.013413	4.736774
5	2015-02-06	0.006623	4.750058	0.043990	0.043990	4.750058	4.750058	4.734311	4.852198	0.059035	0.059035	0.059035	-0.015045	-0.015045	-0.015045	0.043990	4.794048
6	2015-02-07	0.007947	4.749930	0.046731	0.046731	4.749930	4.749930	4.738959	4.854388	0.061440	0.061440	0.061440	-0.014709	-0.014709	-0.014709	0.046731	4.796661
7	2015-02-08	0.009272	4.749801	-0.052159	-0.052159	4.749801	4.749801	4.637375	4.759022	-0.037397	-0.037397	-0.037397	-0.014762	-0.014762	-0.014762	-0.052159	4.697642
8	2015-02-09	0.010596	4.749672	-0.073678	-0.073678	4.749672	4.749672	4.619588	4.733545	-0.058506	-0.058506	-0.058506	-0.015173	-0.015173	-0.015173	-0.073678	4.675994
9	2015-02-10	0.011921	4.749544	-0.046732	-0.046732	4.749544	4.749544	4.649343	4.766111	-0.030829	-0.030829	-0.030829	-0.015903	-0.015903	-0.015903	-0.046732	4.702812
10	2015-02-11	0.013245	4.749415	-0.013031	-0.013031	4.749415	4.749415	4.678914	4.800226	0.003878	0.003878	0.003878	-0.016909	-0.016909	-0.016909	-0.013031	4.736384
11	2015-02-12	0.014570	4.749286	-0.015762	-0.015762	4.749286	4.749286	4.677391	4.795541	0.002379	0.002379	0.002379	-0.018142	-0.018142	-0.018142	-0.015762	4.733524
12	2015-02-14	0.017219	4.749029	0.040359	0.040359	4.749029	4.749029	4.730550	4.852044	0.061440	0.061440	0.061440	-0.021080	-0.021080	-0.021080	0.040359	4.789388
13	2015-02-15	0.018543	4.748900	-0.060076	-0.060076	4.748900	4.748900	4.631445	4.747374	-0.037397	-0.037397	-0.037397	-0.022679	-0.022679	-0.022679	-0.060076	4.688824
14	2015-02-16	0.019868	4.748771	-0.082797	-0.082797	4.748771	4.748771	4.609208	4.724351	-0.058506	-0.058506	-0.058506	-0.024292	-0.024292	-0.024292	-0.082797	4.665974
15	2015-02-17	0.021192	4.748643	-0.056697	-0.056697	4.748643	4.748643	4.637425	4.753577	-0.030829	-0.030829	-0.030829	-0.025868	-0.025868	-0.025868	-0.056697	4.691946
16	2015-02-18	0.022517	4.748514	-0.023482	-0.023482	4.748514	4.748514	4.667891	4.785552	0.003878	0.003878	0.003878	-0.027360	-0.027360	-0.027360	-0.023482	4.725032
17	2015-02-19	0.023841	4.748385	-0.026343	-0.026343	4.748385	4.748385	4.663674	4.783288	0.002379	0.002379	0.002379	-0.028723	-0.028723	-0.028723	-0.026343	4.722042
18	2015-02-20	0.025166	4.748256	0.029117	0.029117	4.748256	4.748256	4.718172	4.833382	0.059035	0.059035	0.059035	-0.029918	-0.029918	-0.029918	0.029117	4.777373
19	2015-02-21	0.026490	4.748128	0.030528	0.030528	4.748128	4.748128	4.716428	4.836553	0.061440	0.061440	0.061440	-0.030911	-0.030911	-0.030911	0.030528	4.778656
20	2015-02-22	0.027815	4.747999	-0.069074	-0.069074	4.747999	4.747999	4.622780	4.741123	-0.037397	-0.037397	-0.037397	-0.031677	-0.031677	-0.031677	-0.069074	4.678925
21	2015-02-23	0.029139	4.747870	-0.090698	-0.090698	4.747870	4.747870	4.599048	4.717505	-0.058506	-0.058506	-0.058506	-0.032193	-0.032193	-0.032193	-0.090698	4.657172
22	2015-02-24	0.030464	4.747742	-0.063274	-0.063274	4.747742	4.747742	4.624169	4.741186	-0.030829	-0.030829	-0.030829	-0.032446	-0.032446	-0.032446	-0.063274	4.684467
23	2015-02-25	0.031788	4.747613	-0.028551	-0.028551	4.747613	4.747613	4.660959	4.778433	0.003878	0.003878	0.003878	-0.032428	-0.032428	-0.032428	-0.028551	4.719062
24	2015-02-26	0.033113	4.747484	-0.029761	-0.029761	4.747484	4.747484	4.658521	4.776376	0.002379	0.002379	0.002379	-0.032141	-0.032141	-0.032141	-0.029761	4.717723
25	2015-02-27	0.034437	4.747355	0.027447	0.027447	4.747355	4.747355	4.717733	4.832807	0.059035	0.059035	0.059035	-0.031588	-0.031588	-0.031588	0.027447	4.774802
26	2015-02-28	0.035762	4.747227	0.030657	0.030657	4.747227	4.747227	4.718431	4.831581	0.061440	0.061440	0.061440	-0.030783	-0.030783	-0.030783	0.030657	4.777883
27	2015-03-01	0.037086	4.747098	-0.067140	-0.067140	4.747098	4.747098	4.620859	4.731811	-0.037397	-0.037397	-0.037397	-0.029743	-0.029743	-0.029743	-0.067140	4.679958
28	2015-03-02	0.038411	4.746969	-0.086995	-0.086995	4.746969	4.746969	4.604208	4.717865	-0.058506	-0.058506	-0.058506	-0.028489	-0.028489	-0.028489	-0.086995	4.659975
29	2015-03-03	0.039735	4.746841	-0.057877	-0.057877	4.746841	4.746841	4.632128	4.747440	-0.030829	-0.030829	-0.030829	-0.027049	-0.027049	-0.027049	-0.057877	4.688963
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
1031	2017-12-29	1.406623	4.708875	0.108752	0.108752	4.692505	4.724074	4.757283	4.872094	0.059035	0.059035	0.059035	0.049717	0.049717	0.049717	0.108752	4.817627
1032	2017-12-30	1.407947	4.708824	0.101238	0.101238	4.692361	4.724086	4.749039	4.871049	0.061440	0.061440	0.061440	0.039798	0.039798	0.039798	0.101238	4.810062
1033	2017-12-31	1.409272	4.708774	-0.007642	-0.007642	4.692231	4.724099	4.643275	4.755815	-0.037397	-0.037397	-0.037397	0.029755	0.029755	0.029755	-0.007642	4.701132
1034	2018-01-01	1.410596	4.708723	-0.038804	-0.038804	4.692108	4.724111	4.612485	4.731552	-0.058506	-0.058506	-0.058506	0.019701	0.019701	0.019701	-0.038804	4.669919
1035	2018-01-02	1.411921	4.708672	-0.021079	-0.021079	4.691985	4.724164	4.630295	4.748048	-0.030829	-0.030829	-0.030829	0.009750	0.009750	0.009750	-0.021079	4.687594
1036	2018-01-03	1.413245	4.708622	0.003890	0.003890	4.691862	4.724231	4.650104	4.770984	0.003878	0.003878	0.003878	0.000013	0.000013	0.000013	0.003890	4.712512
1037	2018-01-04	1.414570	4.708571	-0.007021	-0.007021	4.691733	4.724298	4.640287	4.762276	0.002379	0.002379	0.002379	-0.009401	-0.009401	-0.009401	-0.007021	4.701550
1038	2018-01-05	1.415894	4.708521	0.040647	0.040647	4.691594	4.724316	4.685780	4.809235	0.059035	0.059035	0.059035	-0.018388	-0.018388	-0.018388	0.040647	4.749167
1039	2018-01-06	1.417219	4.708470	0.034587	0.034587	4.691456	4.724363	4.680212	4.807902	0.061440	0.061440	0.061440	-0.026853	-0.026853	-0.026853	0.034587	4.743057
1040	2018-01-07	1.418543	4.708419	-0.072104	-0.072104	4.691300	4.724504	4.570585	4.694350	-0.037397	-0.037397	-0.037397	-0.034707	-0.034707	-0.034707	-0.072104	4.636315
1041	2018-01-08	1.419868	4.708369	-0.100379	-0.100379	4.691060	4.724599	4.548674	4.668777	-0.058506	-0.058506	-0.058506	-0.041874	-0.041874	-0.041874	-0.100379	4.607990
1042	2018-01-09	1.421192	4.708318	-0.079116	-0.079116	4.690914	4.724616	4.568201	4.686485	-0.030829	-0.030829	-0.030829	-0.048288	-0.048288	-0.048288	-0.079116	4.629202
1043	2018-01-10	1.422517	4.708267	-0.050019	-0.050019	4.690769	4.724632	4.593933	4.718391	0.003878	0.003878	0.003878	-0.053897	-0.053897	-0.053897	-0.050019	4.658248
1044	2018-01-11	1.423841	4.708217	-0.056282	-0.056282	4.690624	4.724639	4.589652	4.714692	0.002379	0.002379	0.002379	-0.058661	-0.058661	-0.058661	-0.056282	4.651935
1045	2018-01-12	1.425166	4.708166	-0.003523	-0.003523	4.690480	4.724698	4.644535	4.763789	0.059035	0.059035	0.059035	-0.062558	-0.062558	-0.062558	-0.003523	4.704643
1046	2018-01-13	1.426490	4.708115	-0.004135	-0.004135	4.690335	4.724751	4.648964	4.764602	0.061440	0.061440	0.061440	-0.065575	-0.065575	-0.065575	-0.004135	4.703980
1047	2018-01-14	1.427815	4.708065	-0.105114	-0.105114	4.690191	4.724744	4.541776	4.662461	-0.037397	-0.037397	-0.037397	-0.067717	-0.067717	-0.067717	-0.105114	4.602950
1048	2018-01-15	1.429139	4.708014	-0.127508	-0.127508	4.690049	4.724757	4.521715	4.641925	-0.058506	-0.058506	-0.058506	-0.069003	-0.069003	-0.069003	-0.127508	4.580506
1049	2018-01-16	1.430464	4.707963	-0.100292	-0.100292	4.689906	4.724819	4.543367	4.669104	-0.030829	-0.030829	-0.030829	-0.069463	-0.069463	-0.069463	-0.100292	4.607671
1050	2018-01-17	1.431788	4.707913	-0.065266	-0.065266	4.689776	4.724880	4.580314	4.701597	0.003878	0.003878	0.003878	-0.069144	-0.069144	-0.069144	-0.065266	4.642647
1051	2018-01-18	1.433113	4.707862	-0.065719	-0.065719	4.689654	4.724942	4.584266	4.704747	0.002379	0.002379	0.002379	-0.068099	-0.068099	-0.068099	-0.065719	4.642143
1052	2018-01-19	1.434437	4.707811	-0.007361	-0.007361	4.689527	4.724997	4.638522	4.761000	0.059035	0.059035	0.059035	-0.066396	-0.066396	-0.066396	-0.007361	4.700451
1053	2018-01-20	1.435762	4.707761	-0.002668	-0.002668	4.689399	4.725050	4.644288	4.768587	0.061440	0.061440	0.061440	-0.064108	-0.064108	-0.064108	-0.002668	4.705093
1054	2018-01-21	1.437086	4.707710	-0.098715	-0.098715	4.689273	4.725102	4.544435	4.663375	-0.037397	-0.037397	-0.037397	-0.061318	-0.061318	-0.061318	-0.098715	4.608995
1055	2018-01-22	1.438411	4.707659	-0.116617	-0.116617	4.689109	4.725159	4.529550	4.651315	-0.058506	-0.058506	-0.058506	-0.058112	-0.058112	-0.058112	-0.116617	4.591042
1056	2018-01-23	1.439735	4.707609	-0.085409	-0.085409	4.688953	4.725204	4.560516	4.680854	-0.030829	-0.030829	-0.030829	-0.054580	-0.054580	-0.054580	-0.085409	4.622200
1057	2018-01-24	1.441060	4.707558	-0.046937	-0.046937	4.688895	4.725207	4.596642	4.721010	0.003878	0.003878	0.003878	-0.050815	-0.050815	-0.050815	-0.046937	4.660621
1058	2018-01-25	1.442384	4.707507	-0.044528	-0.044528	4.688741	4.725172	4.599900	4.724595	0.002379	0.002379	0.002379	-0.046907	-0.046907	-0.046907	-0.044528	4.662980
1059	2018-01-26	1.443709	4.707457	0.016089	0.016089	4.688587	4.725120	4.664982	4.786545	0.059035	0.059035	0.059035	-0.042946	-0.042946	-0.042946	0.016089	4.723546
1060	2018-01-27	1.445033	4.707406	0.022423	0.022423	4.688468	4.725110	4.665482	4.793615	0.061440	0.061440	0.061440	-0.039017	-0.039017	-0.039017	0.022423	4.729829

1061 rows × 17 columns

#Get R-squared to measure model variation
stats.corrcoef(x=txns['y'], y=forecast.loc[:724,'yhat'])

array([[ 1.        ,  0.75356776],
       [ 0.75356776,  1.        ]])

Additional information to Model

Scratcher Event & Holidays to consider

events_15 = pd.read_csv('2015_events.csv', parse_dates=['ds'])
events_16 = pd.read_csv('2016_events.csv', parse_dates=['ds'])
events_17 = pd.read_csv('2017_events.csv', parse_dates=['ds'])
events_future = pd.read_csv('2017_events_future.csv', parse_dates=['ds'])
events_future2 = pd.read_csv('2017_events_future2.csv', parse_dates=['ds'])


past_events = events_15.append(events_16)
all_past_events = past_events.append(events_17)
all_events = all_past_events.append(events_future)
all_events2 = all_past_events.append(events_future2)
all_events2

	ds	holiday	lower_window	upper_window
0	2015-03-11	scratch	0	7
1	2015-04-05	holiday	-2	2
2	2015-06-17	scratch	0	7
3	2015-07-04	holiday	-2	2
4	2015-07-15	scratch	0	7
5	2015-08-12	scratch	0	7
6	2015-10-14	scratch	0	7
7	2015-11-11	scratch	0	7
8	2015-11-26	holiday	-2	2
9	2015-12-02	scratch	0	7
10	2015-12-16	scratch	0	7
11	2015-12-25	holiday	-2	2
12	2016-01-01	holiday	-2	2
0	2016-01-27	scratch	0	7
1	2016-03-16	scratch	0	7
2	2016-03-27	holiday	-2	2
3	2016-04-01	scratch	0	7
4	2016-06-17	scratch	0	7
5	2016-07-04	holiday	-2	2
6	2016-07-15	scratch	0	7
7	2016-08-12	scratch	0	7
8	2016-10-14	scratch	0	7
9	2016-11-09	scratch	0	7
10	2016-11-24	holiday	-2	2
11	2016-12-02	scratch	0	7
12	2016-12-14	scratch	0	7
13	2016-12-25	holiday	-2	2
14	2017-01-01	holiday	-2	2
0	2017-02-01	scratch	0	7
1	2017-03-15	scratch	0	7
2	2017-04-16	holiday	-2	2
0	2017-06-28	scratch	0	7
1	2017-07-04	holiday	-2	2
2	2017-08-09	scratch	0	7
3	2017-10-11	scratch	0	7
4	2017-11-10	scratch	0	7
5	2017-11-23	holiday	-2	2
6	2017-11-29	scratch	0	7
7	2017-12-15	scratch	0	7
8	2017-12-25	holiday	-2	2

#read in dates with Scratcher Event & Holidays listed
dates = pd.read_csv('changepoint_holiday.csv',parse_dates=['Scratcher Holiday','Ad Changepoint','Year Changepoint','Holidays'])
#set up different event/holiday dataframes, their 'impact' windows and combine sets
scratch_hol = pd.DataFrame({
    'holiday':'holiday',
    'ds': pd.to_datetime(dates['Scratcher Holiday'].dropna()),
    'lower_window':0,
    'upper_window':7,
})
reg_hol = pd.DataFrame({
    'holiday':'holiday',
    'ds': pd.to_datetime(dates['Holidays'].dropna()),
    'lower_window':-2,
    'upper_window':2,
})
new_dates = pd.DataFrame({
    'ds':pd.to_datetime('2017-11-23'), 
    'holiday':'holiday', 
    'lower_window':-2, 
    'upper_window':2,},
    index=[21])
reg_hol = reg_hol.append(new_dates)

holidays = scratch_hol.append(reg_hol)
holidays.sort( ascending=False)

/home/nbuser/anaconda3_410/lib/python3.5/site-packages/ipykernel/__main__.py:25: FutureWarning: sort(....) is deprecated, use sort_index(.....)

	ds	holiday	lower_window	upper_window
21	2017-11-23	holiday	-2	2
20	2017-04-16	holiday	-2	2
20	2017-06-30	holiday	0	7
19	2017-02-01	holiday	0	7
18	2017-03-15	holiday	0	7
17	2015-03-11	holiday	0	7
16	2016-01-27	holiday	0	7
15	2015-06-17	holiday	0	7
14	2015-07-15	holiday	0	7
13	2015-08-12	holiday	0	7
12	2015-10-14	holiday	0	7
12	2017-12-25	holiday	-2	2
11	2015-11-11	holiday	0	7
11	2017-07-04	holiday	-2	2
10	2015-12-02	holiday	0	7
10	2016-07-04	holiday	-2	2
9	2015-12-16	holiday	0	7
9	2015-07-04	holiday	-2	2
8	2016-03-16	holiday	0	7
8	2017-04-16	holiday	-2	2
7	2016-03-27	holiday	-2	2
7	2016-04-01	holiday	0	7
6	2015-04-05	holiday	-2	2
6	2016-06-17	holiday	0	7
5	2017-01-01	holiday	-2	2
5	2016-07-15	holiday	0	7
4	2016-08-12	holiday	0	7
4	2016-01-01	holiday	-2	2
3	2016-10-14	holiday	0	7
3	2016-12-25	holiday	-2	2
2	2015-12-25	holiday	-2	2
2	2016-11-09	holiday	0	7
1	2016-11-24	holiday	-2	2
1	2016-12-02	holiday	0	7
0	2015-11-26	holiday	-2	2
0	2016-12-14	holiday	0	7

#columns = {'ds', 'holiday', 'lower_window', 'upper_window'}
#rows = '2017-08-11', '2017-10-19', '2017-11-10', '2017-12-01','2017-12-15'

future_events = pd.read_csv('future_events.csv', parse_dates=['ds'])
future_events
#new_holidays = holidays.append(future_events)
#new_holidays

	ds	holiday	lower_window	upper_window
0	2017-08-17	scratch	-7	0
1	2017-10-13	scratch	0	2
2	2017-10-15	scratch	0	2
3	2017-11-10	scratch	0	2
4	2017-11-12	scratch	0	2
5	2017-12-01	scratch	0	2
6	2017-12-03	scratch	0	2
7	2017-12-15	scratch	0	2
8	2017-12-17	scratch	0	2

future_events = pd.read_csv('future_events.csv', parse_dates=['ds'])
future_events
new_holidays = holidays.append(future_events)
new_holidays

	ds	holiday	lower_window	upper_window
0	2016-12-14	holiday	0	7
1	2016-12-02	holiday	0	7
2	2016-11-09	holiday	0	7
3	2016-10-14	holiday	0	7
4	2016-08-12	holiday	0	7
5	2016-07-15	holiday	0	7
6	2016-06-17	holiday	0	7
7	2016-04-01	holiday	0	7
8	2016-03-16	holiday	0	7
9	2015-12-16	holiday	0	7
10	2015-12-02	holiday	0	7
11	2015-11-11	holiday	0	7
12	2015-10-14	holiday	0	7
13	2015-08-12	holiday	0	7
14	2015-07-15	holiday	0	7
15	2015-06-17	holiday	0	7
16	2016-01-27	holiday	0	7
17	2015-03-11	holiday	0	7
18	2017-03-15	holiday	0	7
19	2017-02-01	holiday	0	7
20	2017-06-30	holiday	0	7
0	2015-11-26	holiday	-2	2
1	2016-11-24	holiday	-2	2
2	2015-12-25	holiday	-2	2
3	2016-12-25	holiday	-2	2
4	2016-01-01	holiday	-2	2
5	2017-01-01	holiday	-2	2
6	2015-04-05	holiday	-2	2
7	2016-03-27	holiday	-2	2
8	2017-04-16	holiday	-2	2
9	2015-07-04	holiday	-2	2
10	2016-07-04	holiday	-2	2
11	2017-07-04	holiday	-2	2
12	2017-12-25	holiday	-2	2
20	2017-04-16	holiday	-2	2
21	2017-11-23	holiday	-2	2
0	2017-08-11	holiday	0	7
1	2017-10-13	holiday	0	7
2	2017-11-10	holiday	0	7
3	2017-12-01	holiday	0	7
4	2017-12-15	holiday	0	7

Forecast fitted with Events

m1 = Prophet(holidays=all_events2, holidays_prior_scale=.25)
forecast3 = m1.fit(txns).predict(future)
m1.plot(forecast3);

m1.plot_components(forecast3);

forecast3.to_csv('daily_fcst_all_events2.csv')

forecast1.to_csv('fcst_daily_txns.csv')
txns.to_csv('data_daily_txns.csv')

#Get R-squared to measure model variation
stats.corrcoef(x=txns['y'], y=forecast1.loc[:724,'yhat'])

array([[ 1.        ,  0.79419461],
       [ 0.79419461,  1.        ]])

forecast.to_csv('prophet_txns_fcst.csv')
forecast1.to_csv('prophet_txns_fcst1.csv')

#m1 = Prophet(holidays=holidays, holidays_prior_scale=.25)
forecast1_sales = m1.fit(sales).predict(future)
m1.plot(forecast1_sales);

m1.plot_components(forecast1_sales);

Data by Week

week_data = pd.read_csv('week_data.csv', parse_dates=['ds'])
week_data.describe()

	# Txns	Sales $
count	112.000000	1.120000e+02
mean	387623.285714	2.389304e+07
std	47703.433555	7.113570e+06
min	316991.000000	1.529568e+07
25%	363127.750000	2.034617e+07
50%	375542.500000	2.195671e+07
75%	398826.000000	2.475855e+07
max	596989.000000	5.983885e+07

week_bskt = week_data[['ds','# Txns','Sales $']]
week_bskt['y'] = week_bskt['Sales $']/week_bskt['# Txns']
del week_bskt['Sales $']
del week_bskt['# Txns']
week_bskt

	ds	y
0	2015-02-07	57.751817
1	2015-02-14	60.899264
2	2015-02-21	57.664583
3	2015-02-28	55.852891
4	2015-03-07	54.192692
5	2015-03-14	64.803281
6	2015-03-21	57.812787
7	2015-03-28	54.876293
8	2015-04-04	56.892848
9	2015-04-11	53.913780
10	2015-04-18	62.923289
11	2015-04-25	54.468337
12	2015-05-02	59.306199
13	2015-05-09	57.665705
14	2015-05-16	58.794291
15	2015-05-23	60.335862
16	2015-05-30	58.668581
17	2015-06-06	54.878416
18	2015-06-13	55.465641
19	2015-06-20	68.071180
20	2015-06-27	59.206520
21	2015-07-04	60.860709
22	2015-07-11	53.035462
23	2015-07-18	63.216453
24	2015-07-25	57.996328
25	2015-08-01	57.856493
26	2015-08-08	56.018111
27	2015-08-15	65.600499
28	2015-08-22	60.899654
29	2015-08-29	59.578032
...	...	...
82	2016-09-03	59.212810
83	2016-09-10	57.188839
84	2016-09-17	59.329563
85	2016-09-24	57.261933
86	2016-10-01	57.601069
87	2016-10-08	55.036670
88	2016-10-15	66.931062
89	2016-10-22	60.764614
90	2016-10-29	54.990041
91	2016-11-05	56.560280
92	2016-11-12	67.657786
93	2016-11-19	63.855139
94	2016-11-26	114.508584
95	2016-12-03	71.827240
96	2016-12-10	67.330459
97	2016-12-17	79.066763
98	2016-12-24	83.362573
99	2016-12-31	55.579885
100	2017-01-07	48.252730
101	2017-01-14	53.295858
102	2017-01-21	52.726923
103	2017-01-28	52.179462
104	2017-02-04	59.964624
105	2017-02-11	55.834854
106	2017-02-18	55.241496
107	2017-02-25	59.876513
108	2017-03-04	56.905613
109	2017-03-11	54.436461
110	2017-03-18	65.855111
111	2017-03-25	59.807701

112 rows × 2 columns

week_bskt

	ds	y
0	2015-02-07	57.751817
1	2015-02-14	60.899264
2	2015-02-21	57.664583
3	2015-02-28	55.852891
4	2015-03-07	54.192692
5	2015-03-14	64.803281
6	2015-03-21	57.812787
7	2015-03-28	54.876293
8	2015-04-04	56.892848
9	2015-04-11	53.913780
10	2015-04-18	62.923289
11	2015-04-25	54.468337
12	2015-05-02	59.306199
13	2015-05-09	57.665705
14	2015-05-16	58.794291
15	2015-05-23	60.335862
16	2015-05-30	58.668581
17	2015-06-06	54.878416
18	2015-06-13	55.465641
19	2015-06-20	68.071180
20	2015-06-27	59.206520
21	2015-07-04	60.860709
22	2015-07-11	53.035462
23	2015-07-18	63.216453
24	2015-07-25	57.996328
25	2015-08-01	57.856493
26	2015-08-08	56.018111
27	2015-08-15	65.600499
28	2015-08-22	60.899654
29	2015-08-29	59.578032
...	...	...
82	2016-09-03	59.212810
83	2016-09-10	57.188839
84	2016-09-17	59.329563
85	2016-09-24	57.261933
86	2016-10-01	57.601069
87	2016-10-08	55.036670
88	2016-10-15	66.931062
89	2016-10-22	60.764614
90	2016-10-29	54.990041
91	2016-11-05	56.560280
92	2016-11-12	67.657786
93	2016-11-19	63.855139
94	2016-11-26	114.508584
95	2016-12-03	71.827240
96	2016-12-10	67.330459
97	2016-12-17	79.066763
98	2016-12-24	83.362573
99	2016-12-31	55.579885
100	2017-01-07	48.252730
101	2017-01-14	53.295858
102	2017-01-21	52.726923
103	2017-01-28	52.179462
104	2017-02-04	59.964624
105	2017-02-11	55.834854
106	2017-02-18	55.241496
107	2017-02-25	59.876513
108	2017-03-04	56.905613
109	2017-03-11	54.436461
110	2017-03-18	65.855111
111	2017-03-25	59.807701

112 rows × 2 columns

week_txns = week_data[['ds','# Txns']]
week_txns['y'] = np.log10(week_txns['# Txns'])
del week_txns['# Txns']
week_txns

	ds	y
0	2015-02-07	5.564134
1	2015-02-14	5.638412
2	2015-02-21	5.546383
3	2015-02-28	5.567790
4	2015-03-07	5.562261
5	2015-03-14	5.620316
6	2015-03-21	5.599127
7	2015-03-28	5.572017
8	2015-04-04	5.613450
9	2015-04-11	5.558992
10	2015-04-18	5.608362
11	2015-04-25	5.553200
12	2015-05-02	5.589702
13	2015-05-09	5.627017
14	2015-05-16	5.587549
15	2015-05-23	5.606119
16	2015-05-30	5.573017
17	2015-06-06	5.578356
18	2015-06-13	5.571408
19	2015-06-20	5.617928
20	2015-06-27	5.563689
21	2015-07-04	5.584313
22	2015-07-11	5.532271
23	2015-07-18	5.588061
24	2015-07-25	5.572168
25	2015-08-01	5.569219
26	2015-08-08	5.560179
27	2015-08-15	5.602297
28	2015-08-22	5.582600
29	2015-08-29	5.570953
...	...	...
82	2016-09-03	5.597165
83	2016-09-10	5.544601
84	2016-09-17	5.563593
85	2016-09-24	5.526773
86	2016-10-01	5.560285
87	2016-10-08	5.531421
88	2016-10-15	5.569895
89	2016-10-22	5.569209
90	2016-10-29	5.542246
91	2016-11-05	5.555896
92	2016-11-12	5.576098
93	2016-11-19	5.625310
94	2016-11-26	5.693046
95	2016-12-03	5.668736
96	2016-12-10	5.674914
97	2016-12-17	5.728724
98	2016-12-24	5.775966
99	2016-12-31	5.570002
100	2017-01-07	5.501047
101	2017-01-14	5.533947
102	2017-01-21	5.508561
103	2017-01-28	5.517200
104	2017-02-04	5.563805
105	2017-02-11	5.538898
106	2017-02-18	5.584438
107	2017-02-25	5.549615
108	2017-03-04	5.549207
109	2017-03-11	5.532900
110	2017-03-18	5.580826
111	2017-03-25	5.564596

112 rows × 2 columns

week_txns.to_csv('week_txns.csv')

m_week = Prophet(holidays=new_holidays, changepoints=['2016-01-30','2017-02-04'],changepoint_prior_scale=.25, holidays_prior_scale=.25)

m_week.fit(week_txns)

future_week = m_week.make_future_dataframe(periods=44, freq='W-SAT')

fcst_week = m_week.predict(future_week)
m_week.plot(fcst_week);
#m_week.plot_components(fcst_week);

fcst_week.to_csv('fcst_week.csv')

fcst_week

	ds	t	trend	seasonal_lower	seasonal_upper	trend_lower	trend_upper	yhat_lower	yhat_upper	holiday	holiday_lower	holiday_upper	weekly	weekly_lower	weekly_upper	yearly	yearly_lower	yearly_upper	seasonal	yhat
0	2015-02-07	0.000000	5.550714	0.022272	0.022272	5.550714	5.550714	5.547346	5.602319	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.012477	-0.012477	-0.012477	0.022272	5.572986
1	2015-02-14	0.009009	5.550632	0.025131	0.025131	5.550632	5.550632	5.548403	5.602419	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.009617	-0.009617	-0.009617	0.025131	5.575763
2	2015-02-21	0.018018	5.550550	0.020610	0.020610	5.550550	5.550550	5.544827	5.596707	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.014138	-0.014138	-0.014138	0.020610	5.571160
3	2015-02-28	0.027027	5.550468	0.015380	0.015380	5.550468	5.550468	5.536733	5.593043	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.019369	-0.019369	-0.019369	0.015380	5.565848
4	2015-03-07	0.036036	5.550386	0.013072	0.013072	5.550386	5.550386	5.534867	5.591884	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.021677	-0.021677	-0.021677	0.013072	5.563457
5	2015-03-14	0.045045	5.550304	0.037759	0.037759	5.550304	5.550304	5.559119	5.615523	0.022564	0.022564	0.022564	0.034748	0.034748	0.034748	-0.019554	-0.019554	-0.019554	0.037759	5.588062
6	2015-03-21	0.054054	5.550221	0.021369	0.021369	5.550221	5.550221	5.545527	5.600468	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.013379	-0.013379	-0.013379	0.021369	5.571591
7	2015-03-28	0.063063	5.550139	0.028211	0.028211	5.550139	5.550139	5.552033	5.605434	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.006537	-0.006537	-0.006537	0.028211	5.578350
8	2015-04-04	0.072072	5.550057	0.059433	0.059433	5.550057	5.550057	5.583734	5.636675	0.028293	0.028293	0.028293	0.034748	0.034748	0.034748	-0.003609	-0.003609	-0.003609	0.059433	5.609490
9	2015-04-11	0.081081	5.549975	0.029656	0.029656	5.549975	5.549975	5.552579	5.606949	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.005093	-0.005093	-0.005093	0.029656	5.579631
10	2015-04-18	0.090090	5.549893	0.029618	0.029618	5.549893	5.549893	5.552197	5.606802	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.005131	-0.005131	-0.005131	0.029618	5.579511
11	2015-04-25	0.099099	5.549811	0.037481	0.037481	5.549811	5.549811	5.561058	5.616367	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	0.002732	0.002732	0.002732	0.037481	5.587292
12	2015-05-02	0.108108	5.549729	0.051648	0.051648	5.549729	5.549729	5.575196	5.627618	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	0.016899	0.016899	0.016899	0.051648	5.601376
13	2015-05-09	0.117117	5.549647	0.061801	0.061801	5.549647	5.549647	5.583488	5.637936	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	0.027053	0.027053	0.027053	0.061801	5.611448
14	2015-05-16	0.126126	5.549565	0.058984	0.058984	5.549565	5.549565	5.581563	5.635521	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	0.024236	0.024236	0.024236	0.058984	5.608549
15	2015-05-23	0.135135	5.549482	0.045775	0.045775	5.549482	5.549482	5.568837	5.622621	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	0.011027	0.011027	0.011027	0.045775	5.595258
16	2015-05-30	0.144144	5.549400	0.034229	0.034229	5.549400	5.549400	5.554700	5.611318	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.000519	-0.000519	-0.000519	0.034229	5.583630
17	2015-06-06	0.153153	5.549318	0.033000	0.033000	5.549318	5.549318	5.553562	5.610073	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.001748	-0.001748	-0.001748	0.033000	5.582318
18	2015-06-13	0.162162	5.549236	0.038213	0.038213	5.549236	5.549236	5.560328	5.615115	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	0.003464	0.003464	0.003464	0.038213	5.587449
19	2015-06-20	0.171171	5.549154	0.061317	0.061317	5.549154	5.549154	5.583404	5.636378	0.022564	0.022564	0.022564	0.034748	0.034748	0.034748	0.004004	0.004004	0.004004	0.061317	5.610471
20	2015-06-27	0.180180	5.549072	0.029693	0.029693	5.549072	5.549072	5.551153	5.606128	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.005056	-0.005056	-0.005056	0.029693	5.578764
21	2015-07-04	0.189189	5.548990	0.031744	0.031744	5.548990	5.548990	5.553790	5.607454	0.013767	0.013767	0.013767	0.034748	0.034748	0.034748	-0.016771	-0.016771	-0.016771	0.031744	5.580734
22	2015-07-11	0.198198	5.548908	0.014009	0.014009	5.548908	5.548908	5.533310	5.590121	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.020740	-0.020740	-0.020740	0.014009	5.562916
23	2015-07-18	0.207207	5.548825	0.042311	0.042311	5.548825	5.548825	5.563782	5.619592	0.022564	0.022564	0.022564	0.034748	0.034748	0.034748	-0.015002	-0.015002	-0.015002	0.042311	5.591136
24	2015-07-25	0.216216	5.548743	0.027551	0.027551	5.548743	5.548743	5.548741	5.603111	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.007198	-0.007198	-0.007198	0.027551	5.576294
25	2015-08-01	0.225225	5.548661	0.030252	0.030252	5.548661	5.548661	5.551049	5.605165	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.004497	-0.004497	-0.004497	0.030252	5.578913
26	2015-08-08	0.234234	5.548579	0.029247	0.029247	5.548579	5.548579	5.549797	5.606171	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.005502	-0.005502	-0.005502	0.029247	5.577826
27	2015-08-15	0.243243	5.548497	0.053315	0.053315	5.548497	5.548497	5.575350	5.629704	0.022564	0.022564	0.022564	0.034748	0.034748	0.034748	-0.003997	-0.003997	-0.003997	0.053315	5.601812
28	2015-08-22	0.252252	5.548415	0.035880	0.035880	5.548415	5.548415	5.557566	5.610936	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	0.001132	0.001132	0.001132	0.035880	5.584295
29	2015-08-29	0.261261	5.548333	0.037962	0.037962	5.548333	5.548333	5.558949	5.614455	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	0.003214	0.003214	0.003214	0.037962	5.586295
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
126	2017-07-08	1.135135	5.531225	0.014153	0.014153	5.531189	5.531225	5.517719	5.572352	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.020595	-0.020595	-0.020595	0.014153	5.545378
127	2017-07-15	1.144144	5.531021	0.016946	0.016946	5.530965	5.531021	5.520654	5.575053	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.017802	-0.017802	-0.017802	0.016946	5.547967
128	2017-07-22	1.153153	5.530817	0.025123	0.025123	5.530733	5.530830	5.527898	5.582792	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.009625	-0.009625	-0.009625	0.025123	5.555939
129	2017-07-29	1.162162	5.530613	0.029981	0.029981	5.530510	5.530650	5.533979	5.588919	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.004768	-0.004768	-0.004768	0.029981	5.560593
130	2017-08-05	1.171171	5.530408	0.029626	0.029626	5.530280	5.530466	5.532369	5.586185	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.005122	-0.005122	-0.005122	0.029626	5.560034
131	2017-08-12	1.180180	5.530204	0.055365	0.055365	5.530051	5.530290	5.558192	5.613456	0.025705	0.025705	0.025705	0.034748	0.034748	0.034748	-0.005089	-0.005089	-0.005089	0.055365	5.585569
132	2017-08-19	1.189189	5.530000	0.033945	0.033945	5.529821	5.530128	5.536884	5.591282	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.000804	-0.000804	-0.000804	0.033945	5.563945
133	2017-08-26	1.198198	5.529796	0.038098	0.038098	5.529588	5.529965	5.539165	5.595154	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	0.003350	0.003350	0.003350	0.038098	5.567895
134	2017-09-02	1.207207	5.529592	0.034461	0.034461	5.529348	5.529788	5.537397	5.591977	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.000287	-0.000287	-0.000287	0.034461	5.564053
135	2017-09-09	1.216216	5.529388	0.022198	0.022198	5.529072	5.529622	5.524708	5.579176	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.012550	-0.012550	-0.012550	0.022198	5.551586
136	2017-09-16	1.225225	5.529184	0.009439	0.009439	5.528827	5.529481	5.509105	5.564325	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.025309	-0.025309	-0.025309	0.009439	5.538624
137	2017-09-23	1.234234	5.528980	0.004531	0.004531	5.528579	5.529348	5.505361	5.560463	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.030218	-0.030218	-0.030218	0.004531	5.533511
138	2017-09-30	1.243243	5.528776	0.007305	0.007305	5.528346	5.529180	5.507304	5.562332	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.027444	-0.027444	-0.027444	0.007305	5.536081
139	2017-10-07	1.252252	5.528572	0.010963	0.010963	5.528111	5.529007	5.511260	5.567053	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.023785	-0.023785	-0.023785	0.010963	5.539535
140	2017-10-14	1.261261	5.528368	0.037072	0.037072	5.527878	5.528861	5.536179	5.590811	0.025705	0.025705	0.025705	0.034748	0.034748	0.034748	-0.023382	-0.023382	-0.023382	0.037072	5.565440
141	2017-10-21	1.270270	5.528164	0.011433	0.011433	5.527615	5.528707	5.511572	5.566162	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.023315	-0.023315	-0.023315	0.011433	5.539598
142	2017-10-28	1.279279	5.527960	0.015669	0.015669	5.527335	5.528554	5.516570	5.570857	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.019079	-0.019079	-0.019079	0.015669	5.543629
143	2017-11-04	1.288288	5.527756	0.023588	0.023588	5.527046	5.528426	5.525306	5.577934	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.011161	-0.011161	-0.011161	0.023588	5.551344
144	2017-11-11	1.297297	5.527552	0.057902	0.057902	5.526761	5.528291	5.558400	5.611796	0.025705	0.025705	0.025705	0.034748	0.034748	0.034748	-0.002552	-0.002552	-0.002552	0.057902	5.585454
145	2017-11-18	1.306306	5.527348	0.044324	0.044324	5.526522	5.528191	5.543959	5.598125	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	0.009576	0.009576	0.009576	0.044324	5.571672
146	2017-11-25	1.315315	5.527144	0.143716	0.143716	5.526262	5.528090	5.643038	5.699013	0.074386	0.074386	0.074386	0.034748	0.034748	0.034748	0.034582	0.034582	0.034582	0.143716	5.670860
147	2017-12-02	1.324324	5.526940	0.136539	0.136539	5.526010	5.527945	5.636818	5.691185	0.025705	0.025705	0.025705	0.034748	0.034748	0.034748	0.076086	0.076086	0.076086	0.136539	5.663479
148	2017-12-09	1.333333	5.526736	0.154170	0.154170	5.525773	5.527800	5.652723	5.708267	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	0.119422	0.119422	0.119422	0.154170	5.680906
149	2017-12-16	1.342342	5.526532	0.196849	0.196849	5.525507	5.527647	5.697632	5.750509	0.025705	0.025705	0.025705	0.034748	0.034748	0.034748	0.136396	0.136396	0.136396	0.196849	5.723381
150	2017-12-23	1.351351	5.526328	0.203197	0.203197	5.525218	5.527499	5.700713	5.756401	0.060733	0.060733	0.060733	0.034748	0.034748	0.034748	0.107715	0.107715	0.107715	0.203197	5.729525
151	2017-12-30	1.360360	5.526124	0.077227	0.077227	5.524898	5.527342	5.575491	5.630344	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	0.042479	0.042479	0.042479	0.077227	5.603351
152	2018-01-06	1.369369	5.525920	0.008948	0.008948	5.524630	5.527201	5.507735	5.562164	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.025800	-0.025800	-0.025800	0.008948	5.534868
153	2018-01-13	1.378378	5.525716	-0.029610	-0.029610	5.524361	5.527062	5.468948	5.522439	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.064358	-0.064358	-0.064358	-0.029610	5.496106
154	2018-01-20	1.387387	5.525512	-0.029530	-0.029530	5.524081	5.526952	5.469103	5.521989	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.064278	-0.064278	-0.064278	-0.029530	5.495982
155	2018-01-27	1.396396	5.525308	-0.006863	-0.006863	5.523752	5.526792	5.491214	5.546574	0.000000	0.000000	0.000000	0.034748	0.034748	0.034748	-0.041611	-0.041611	-0.041611	-0.006863	5.518445

156 rows × 20 columns

#Get R-squared to measure model variation
stats.corrcoef(x=week_txns['y'], y=fcst_week.loc[:111,'yhat'])

week_sales = week_data[['ds','Sales $']]
week_sales['y'] = np.log10(week_sales['Sales $'])
del week_sales['Sales $']
week_sales

m_bskt = Prophet(holidays=all_events)
m_bskt.fit(week_bskt)
fcst_week_bskt = m_bskt.predict(future_week)
m_bskt.plot(fcst_week_bskt);

fcst_week_bskt.to_csv('fcst_week_bskt_all_events.csv')

m_week.plot_components(fcst_week_sales);
#m_week.yearly_seasonality(fcst_week_sales)

m_week.plot_components(fcst_week_sales);

fcst_week
fcst_week.to_csv('fcst_week_sales.csv')
week_sales.to_csv('data_week_sales.csv')

#Get R-squared to measure model variation
stats.corrcoef(x=week_txns['y'], y=fcst_week.loc[:111,'yhat'])

m_week.fit(week_bskt)
fcst_week_bskt = m_week.predict(future_week)
m_week.plot(fcst_week_bskt);

fcst_week_bskt.to_csv('fcst_week_bskt.csv')

Data by GMM

mg1_data = pd.read_csv('mg1_week_data.csv', parse_dates=['ds'])
mg2_data = pd.read_csv('mg2_week_data.csv', parse_dates=['ds'])
mg3_data = pd.read_csv('mg3_week_data.csv', parse_dates=['ds'])
#mg1_data.info()
#mg2_data.describe()
mg3_data.describe()

#mg1 = pd.DataFrame(mg1_data[['ds','Txns']])
#mg1['y'] = np.log10(mg1['Txns'])
#del mg1['Txns']
#mg1.drop(mg1.index[-1],inplace= True)
mg1

#mg2 = pd.DataFrame(mg2_data[['ds','Txns']])
#mg2['y'] = np.log10(mg2['Txns'])
#del mg2['Txns']
#mg2.drop(mg2.index[-1],inplace= True)
mg2

#mg3 = pd.DataFrame(mg3_data[['ds','Txns']])
#mg3['y'] = np.log10(mg3['Txns'])
#del mg3['Txns']
#mg3.drop(mg3.index[-1],inplace= True)
mg3

m_week.fit(mg1)
fcst_mg1_txns = m_week.predict(future_week)
fcst_mg1_txns

m_week.plot(fcst_mg1_txns);

m_week.plot_components(fcst_mg1_txns);

m_week.fit(mg2)
fcst_mg2_txns = m_week.predict(future_week)
fcst_mg2_txns

m_week.plot(fcst_mg2_txns);

m_week.plot_components(fcst_mg2_txns);

m_week.fit(mg3)
fcst_mg3_txns = m_week.predict(future_week)
fcst_mg3_txns

m_week.plot(fcst_mg3_txns);

m_week.plot_components(fcst_mg3_txns);