Working as a Data Analyst - Method

As previously mentioned, the realm of data work is dynamic and multifaceted, entailing the integration of numerous components to execute a complete process. Consequently, it becomes pivotal to devise an effective approach to manage these complexities.

Certain tasks within this domain may appear monotonous. For instance, in the course of my duties, the data procured via sensors often adopts unconventional formats, deviating from the standard CSV or TXT formats. This necessitates the conversion of such formats into a comprehensible structure. Furthermore, the data obtained today could differ markedly in format and nomenclature from that acquired years down the line. Alternatively, there are instances when swift calculations of statistical metrics from arrays are imperative.

Although these tasks might appear routine and uninteresting, they form the bedrock of ensuring the data is refined, validated, and primed for utilization prior to embarking on any sophisticated modelling endeavours.

On the other hand, sometimes work could be abstract and open-ended. “路漫漫其修远兮，吾将上下而求索”, The path of research is one that seldom offers a clear endpoint. It's therefore of paramount importance to maintain composure and persist steadfastly. As a student in the field of material engineering, I've come to understand that exploratory endeavours demand time and can be exasperating, as evident in the realm of superconductors.

The same principle applies within the realm of data analysis. There are instances when it's simply unfeasible to unearth a flawless approach to dissecting data and producing impeccable outcomes. Take, for instance, our extensive efforts in exploring features derived from both the time and frequency domains. Regrettably, only a scant few have displayed meaningful correlations with our desired objectives.

In light of this, the necessity arises to innovate and uncover novel methodologies to unveil the intrinsic connections between quantifiable attributes and the ultimate target. This process parallels the dynamic nature of research, wherein perseverance and the pursuit of new avenues remain pivotal.

Anyway, I have mentioned two general types of work that could be met during data work. When it comes to the machining type of work, it is important to keep the work simple and straightforward. I personally like to ask my supervisor and myself 2 questions, What is input and what is output? After I know those questions and find the best methods to do it, I will try to concentrate and finish it as soon as possible. However, when it comes to explorative work, it is important to keep the options open and never get caught up in it. Sometimes, I get obsessed with one method and waste too much time on proving it works. I would like to use the method I used when I was in high school my teacher in China taught me about writing a paper. “When you start to write your paper, you should look at the title of the thesis in the end, let your thought fly for a while”.

In conclusion, to keep yourself efficient, it is important to understand the type of work you are doing and choose according method to finish it.