First Derivative of the X values vs Y in line chart

[PBIBee321]

Hi All,

I want to plot the first derivative d(I)/d(V) as my Y axis and V as my X axis.The below is my sample table .I am not sure how can I plot the first derivative of d(I)/d(v) vs V in a line chart. Is there any measure we can write to handle this?

I tried in excel but it didnt work (A4-A2)/(B4-B2) assuming A as V and B as I column. I got DIV error in excel. I am now trying in PBI, using any measures or calculated columns  approach to calculate the first derivative for my V values ,can someone provide an approach? 

My data source table is as below:-

V	I
0.007	2.157
0.015	2.279
0.023	2.279
0.031	2.206
0.039	2.328
0.047	2.304
0.055	2.328
0.063	2.426
0.071	2.377
0.079	2.426
0.086	2.426
0.094	2.377
0.102	2.475
0.11	2.451
0.118	2.451
0.126	2.5
0.134	2.551
0.142	2.574
0.15	2.574
0.158	2.549
0.166	2.574
0.173	2.549
0.181	2.623
0.189	2.672
0.197	2.623
0.205	2.696
0.213	2.721
0.221	2.647
0.229	2.745
0.237	2.696
0.245	2.77
0.252	2.819
0.26	2.77
0.268	2.843
0.276	2.868
0.284	2.819
0.292	2.892
0.3	2.843
0.308	2.892
0.316	2.941
0.324	2.868
0.332	2.941
0.339	2.99
0.347	2.917
0.355	2.99
0.363	2.941
0.371	3.015
0.379	3.015
0.387	2.966
0.395	3.015
0.403	3.088
0.411	3.015
0.418	3.088
0.426	3.015
0.434	3.088
0.442	3.113
0.45	3.039
0.458	3.088
0.466	3.113
0.474	3.064
0.482	3.137
0.49	3.064
0.497	3.137
0.505	3.137
0.513	3.088
0.521	3.137
0.529	3.162
0.537	3.113
0.545	3.162
0.553	3.113
0.561	3.137
0.569	3.186
0.577	3.137
0.584	3.162
0.592	3.211
0.6	3.137
0.608	3.186
0.616	3.137
0.624	3.186
0.632	3.211
0.64	3.137
0.648	3.211
0.656	3.235
0.663	3.162
0.671	3.211
0.679	3.162
0.687	3.211
0.695	3.235
0.703	3.186
0.711	3.235
0.719	3.26
0.727	3.186
0.735	3.235
0.743	3.186
0.75	3.235
0.758	3.26
0.766	3.186
0.774	3.235
0.782	3.26
0.79	3.211
0.798	3.26
0.806	3.186
0.814	3.235
0.822	3.26
0.829	3.211
0.837	3.26
0.845	3.284
0.853	3.211
0.861	3.26
0.869	3.211
0.877	3.26
0.885	3.284
0.893	3.211
0.901	3.26
0.908	3.284
0.916	3.211
0.924	3.26
0.932	3.211
0.94	3.26
0.948	3.284
0.956	3.211
0.964	3.26
0.972	3.309
0.98	3.235
0.988	3.284
0.995	3.211
1.004	3.26
1.012	3.284
1.02	3.235
1.027	3.284
1.035	3.309
1.043	3.235
1.051	3.284
1.059	3.235
1.067	3.284
1.075	3.309
1.083	3.235
1.091	3.284
1.099	3.309
1.107	3.235
1.114	3.309
1.122	3.235
1.13	3.284
1.138	3.309
1.146	3.26
1.154	3.309
1.162	3.333
1.17	3.26
1.178	3.309
1.186	3.26
1.193	3.309
1.201	3.333
1.209	3.26
1.217	3.333
1.225	3.333
1.233	3.284
1.241	3.333
1.249	3.284
1.257	3.333
1.265	3.358
1.273	3.284
1.28	3.358
1.288	3.382
1.296	3.309
1.304	3.358
1.312	3.284
1.32	3.358
1.328	3.382
1.336	3.309
1.344	3.382
1.352	3.407
1.359	3.333
1.367	3.382
1.375	3.333
1.383	3.382
1.391	3.407
1.399	3.333
1.407	3.407
1.415	3.431
1.423	3.358
1.431	3.407
1.438	3.358
1.446	3.407
1.454	3.431
1.462	3.382
1.47	3.431
1.478	3.456
1.486	3.382
1.494	3.456
1.502	3.382
1.51	3.456
1.518	3.48
1.525	3.431
1.533	3.48
1.541	3.505
1.549	3.431
1.557	3.505
1.565	3.431
1.573	3.505
1.581	3.529
1.589	3.456
1.597	3.529
1.604	3.554
1.612	3.48
1.62	3.529
1.628	3.48
1.636	3.529
1.644	3.578
1.652	3.505
1.66	3.578
1.668	3.627
1.676	3.529
1.684	3.578
1.691	3.529
1.699	3.578
1.707	3.603
1.715	3.554
1.723	3.603
1.731	3.652
1.739	3.554
1.747	3.627
1.755	3.554
1.763	3.627
1.77	3.676
1.778	3.627
1.786	3.676
1.794	3.701
1.802	3.627
1.81	3.701
1.818	3.603
1.826	3.676
1.834	3.725
1.842	3.652
1.849	3.725
1.857	3.75
1.865	3.676
1.873	3.75
1.881	3.676
1.889	3.75
1.897	3.775
1.905	3.725
1.913	3.799
1.921	3.824
1.929	3.75
1.936	3.799
1.944	3.775
1.952	3.824
1.96	3.848
1.968	3.799
1.976	3.848
1.984	3.897
1.992	3.848
2	3.946
2.008	3.873
2.015	3.946

Expected dI/dV(as Y axis) vs V (As X axis)plot

I know my values are not much tidy, I just want to know the kind of plot I will get for my above values.Even if it is not close to the expected screenshot below

Kind regards

Guthu

[AlexisOlson]

I'd strongly recommend smoothing your data before trying to do anything like a derivative. In this particular case, you can approximate your data quite well with a 4th-degree polynomial and then take the derivative of that approximation.

This will lead to much cleaner results like this:

[AlexisOlson · Accepted solution]

Your data is pretty noisy, but you can still do the difference quotient approximation as follows:

dI/dV =
VAR V_Curr = SELECTEDVALUE ( Table1[V] )
VAR V_Prev = CALCULATE ( MAX ( Table1[V] ), Table1[V] < V_Curr )
VAR I_Curr = CALCULATE ( SUM ( Table1[I] ), Table1[V] = V_Curr )
VAR I_Prev = CALCULATE ( SUM ( Table1[I] ), Table1[V] = V_Prev )
RETURN
    DIVIDE ( I_Curr - I_Prev, V_Curr - V_Prev )

Since your data is so noisy, this is all over the place: