Scaling data with heavier weight near zero [Hysteresis]

[AdventureMike]

Magnetic data taking is done with preference to the origin of the graph. More points are taken near zero x-values than at the end points. I want to expand the middle of a graph in a sensible way to show the important data at the middle while also having the end points in frame. I have tried to follow this beautiful post: "peltiertech reciprocal-chart-axis-scale" but there are some difficulties in that my data spans the four quadrants of the graph and is not one to one in that an important part of it is the non-overlap of the central area.

My ideal x-axis divisions are: ~-70000, -50000, -30000, -20000, -10000, -9000, -8000, ..., 1000, 2000, ..., 10000, 20000, 30000, 50000, 70000
Anything that is near those ratios is fine.
Attached is one example of my data.

Thank you fellow number and graph lovers.

[shg]

You could do a non-linear transformation like arctangent:

	A	B	C	D	E	F
1	x	y		70000.4	1.27E-05	D1: {=MAX(ABS(A3:A83))}
2	Magnetic Field (Oe)	ssx 1.8k
3	70000.3359	12017E-9		1.0000	0.9634	D3: =DEGREES(ATAN(A3/D$1)/45)
4	63175.2344	12616E-9		0.9348	0.9944
5	56700.3750	12634E-9		0.8668	0.9953
6	50575.2852	12728E-9		0.7966	1.0000
7	44800.1758	12143E-9		0.7249	0.9700
8	39375.3945	12022E-9		0.6524	0.9637
9	34300.4453	11214E-9		0.5801	0.9196
10	29575.3691	10371E-9		0.5090	0.8705
11	25200.3828	9567E-9		0.4400	0.8207
12	21175.2344	8627E-9		0.3740	0.7584
13	17500.2090	7612E-9		0.3119	0.6863
14	14175.4785	6419E-9		0.2544	0.5948
15	11200.2217	5285E-9		0.2020	0.5011
16	8575.1846	4167E-9		0.1552	0.4028

You'd need to add other series with data labels for axis values.

[shg]

Or take a root:

	A	B	C	D	E	F
1	x	y		2
2	Magnetic Field (Oe)	ssx 1.8k		264.5759	0.0036	D2: {=MAX(ABS(A3:A83))^(1/$D$1)}
3	70000.336	12017E-9		1.0000	0.9717	D3: =SIGN(A3)*ABS(A3)^(1/$D$1)/D$2
4	63175.234	12616E-9		0.9500	0.9956
5	56700.375	12634E-9		0.9000	0.9963
6	50575.285	12728E-9		0.8500	1.0000
7	44800.176	12143E-9		0.8000	0.9767
8	39375.395	12022E-9		0.7500	0.9719
9	34300.445	11214E-9		0.7000	0.9386
10	29575.369	10371E-9		0.6500	0.9027
11	25200.383	9567E-9		0.6000	0.8670
12	21175.234	8627E-9		0.5500	0.8233
13	17500.209	7612E-9		0.5000	0.7733
14	14175.479	6419E-9		0.4500	0.7102
15	11200.222	5285E-9		0.4000	0.6444
16	8575.185	4167E-9		0.3500	0.5722

[AdventureMike]

Thank you shg. I see that arc tangent gets me a cluster of data points near the center, but I am unsure on the further scaling of the graph using that.
I'm uploading what I have done. It shows the beautiful gridlines but not the hoped for zoom on the center of the data...
Also, looking at my zoom factor in the called out x-axis values in my first post I'd prefer to have a better look at the hundreds values near the origin. ( ~-70000, -50000, -30000, -20000, -10000, -8000, -6000, ...,-1000, -900, -800, -700,...0, 100, 200, ..., 1000, 2000, ..., 10000, 20000, 30000, 50000, 70000) with even divisions of the grid lines between there.
I feel close to "aha" but may need another little push.

[shg]

Try the second one.

And a smooth line plot makes it easier to see the hysteresis.

[shg]

See attached.

[AdventureMike]

This is damned near perfect. Thanks shg. I am getting the same graph as you showed and am now trying to make a scale that doesn't look too childish.
Really a heart-felt thanks for the support.

[shg]

You're welcome.

[shg]

Someone else may find this useful ...